amphora commited on
Commit
42c1815
·
verified ·
1 Parent(s): 66cd5c8

v2: add 216 optimize tasks / 77 families (Tao constants, AlphaEvolve, Leaps in Bounds, FrontierShift, FrontierCO MDS, ALE-Bench, FrontierCS, EdgeBench)

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. LICENSE.md +5 -3
  2. README.md +5 -3
  3. families.csv +220 -143
  4. tasks/optimize-ae-p1-ff-nikodym-2d-q121/environment/Dockerfile +7 -0
  5. tasks/optimize-ae-p1-ff-nikodym-2d-q121/environment/instance.json +1 -0
  6. tasks/optimize-ae-p1-ff-nikodym-2d-q121/instruction.md +21 -0
  7. tasks/optimize-ae-p1-ff-nikodym-2d-q121/solution/solve.sh +5 -0
  8. tasks/optimize-ae-p1-ff-nikodym-2d-q121/task.toml +43 -0
  9. tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/check.py +57 -0
  10. tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/instance.json +1 -0
  11. tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/run.py +127 -0
  12. tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/test.sh +9 -0
  13. tasks/optimize-ae-p1-ff-nikodym-2d-q25/environment/Dockerfile +7 -0
  14. tasks/optimize-ae-p1-ff-nikodym-2d-q25/environment/instance.json +1 -0
  15. tasks/optimize-ae-p1-ff-nikodym-2d-q25/instruction.md +21 -0
  16. tasks/optimize-ae-p1-ff-nikodym-2d-q25/solution/solve.sh +5 -0
  17. tasks/optimize-ae-p1-ff-nikodym-2d-q25/task.toml +43 -0
  18. tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/check.py +57 -0
  19. tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/instance.json +1 -0
  20. tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/run.py +127 -0
  21. tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/test.sh +9 -0
  22. tasks/optimize-ae-p1-ff-nikodym-2d-q49/environment/Dockerfile +7 -0
  23. tasks/optimize-ae-p1-ff-nikodym-2d-q49/environment/instance.json +1 -0
  24. tasks/optimize-ae-p1-ff-nikodym-2d-q49/instruction.md +21 -0
  25. tasks/optimize-ae-p1-ff-nikodym-2d-q49/solution/solve.sh +5 -0
  26. tasks/optimize-ae-p1-ff-nikodym-2d-q49/task.toml +43 -0
  27. tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/check.py +57 -0
  28. tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/instance.json +1 -0
  29. tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/run.py +127 -0
  30. tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/test.sh +9 -0
  31. tasks/optimize-ae-p14-hausdorff-young-p4-3/environment/Dockerfile +7 -0
  32. tasks/optimize-ae-p14-hausdorff-young-p4-3/environment/instance.json +1 -0
  33. tasks/optimize-ae-p14-hausdorff-young-p4-3/instruction.md +20 -0
  34. tasks/optimize-ae-p14-hausdorff-young-p4-3/solution/solve.sh +0 -0
  35. tasks/optimize-ae-p14-hausdorff-young-p4-3/task.toml +43 -0
  36. tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/check.py +42 -0
  37. tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/instance.json +1 -0
  38. tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/run.py +127 -0
  39. tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/test.sh +9 -0
  40. tasks/optimize-ae-p14-hausdorff-young-p6-5/environment/Dockerfile +7 -0
  41. tasks/optimize-ae-p14-hausdorff-young-p6-5/environment/instance.json +1 -0
  42. tasks/optimize-ae-p14-hausdorff-young-p6-5/instruction.md +20 -0
  43. tasks/optimize-ae-p14-hausdorff-young-p6-5/solution/solve.sh +0 -0
  44. tasks/optimize-ae-p14-hausdorff-young-p6-5/task.toml +43 -0
  45. tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/check.py +42 -0
  46. tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/instance.json +1 -0
  47. tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/run.py +127 -0
  48. tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/test.sh +9 -0
  49. tasks/optimize-ae-p33-thomson-n292/environment/Dockerfile +7 -0
  50. tasks/optimize-ae-p33-thomson-n292/environment/instance.json +1 -0
LICENSE.md CHANGED
@@ -2,8 +2,10 @@
2
 
3
  This collection repackages tasks from several sources. Each row's `license` column gives the license of its source:
4
 
5
- - MIT: RLVE-Gym, NPPC, EinsteinArena (platform and baselines)
6
- - CC-BY-4.0: MathConstraint; AlphaEvolve repository of problems (materials)
7
- - Apache-2.0: Reasoning Gym, FunSearch, AlphaEvolve repository of problems (code), Finch / Evolution Fine-Tuning (code)
 
 
8
 
9
  Please attribute the original authors when using the corresponding tasks.
 
2
 
3
  This collection repackages tasks from several sources. Each row's `license` column gives the license of its source:
4
 
5
+ - MIT: RLVE-Gym, NPPC, EinsteinArena (platform and baselines), Leaps in Bounds (D. Munro), FrontierCO, FrontierCS
6
+ - CC-BY-4.0: MathConstraint; AlphaEvolve repository of problems (materials); EdgeBench (task directories; the underlying AtCoder problem statements remain © AtCoder)
7
+ - Apache-2.0: Reasoning Gym, FunSearch, AlphaEvolve repository of problems (code), Finch / Evolution Fine-Tuning (code), Optimization Constants in Mathematics (T. Tao, D. Davis, P. Ivanisvili et al.)
8
+ - **CC-BY-ND-4.0 (restricted): ALE-Bench (Sakana AI; AtCoder Heuristic Contest problems).** No derivatives: statements are reproduced verbatim; do not redistribute modified versions of these tasks without checking compatibility. Tagged `license_restricted`.
9
+ - **unknown: FrontierShift open problems (AItonomy Foundation).** The public registry states no license and its repository is not public. Tagged `license_unknown`.
10
 
11
  Please attribute the original authors when using the corresponding tasks.
README.md CHANGED
@@ -1,6 +1,6 @@
1
  ---
2
  license: other
3
- license_name: mixed-permissive
4
  license_link: LICENSE.md
5
  tags:
6
  - math
@@ -13,7 +13,7 @@ size_categories:
13
 
14
  # MathConstructOptimize-Envs (Harbor)
15
 
16
- The 3,577 tasks of [`amphora/MathConstructOptimize-Envs`](https://huggingface.co/datasets/amphora/MathConstructOptimize-Envs) as runnable [Harbor](https://github.com/laude-institute/harbor) environments. Each task is one directory:
17
 
18
  ```
19
  tasks/<task_id>/
@@ -34,8 +34,10 @@ harbor run -p mcoe/tasks -a terminus-2 -m <model> -n 16
34
  ```
35
 
36
  **Validation:**
37
- - The oracle agent passes on all 3,577 tasks (construct = 1.0, optimize ≥ 0.1), and a no-op agent scores 0.
38
  - An HTTP canary confirmed that `no-network` blocks internet access inside the task container. Research-level best constructions are public, so do not relax this for RL.
39
  - If you run many trials concurrently, raise Docker's default address pools. Each no-network task creates its own networks, and the defaults are exhausted at roughly 30 concurrent trials.
40
 
 
 
41
  See the main dataset card for the reward contract, tags, sources and the upstream verifier bugs we fixed.
 
1
  ---
2
  license: other
3
+ license_name: mixed
4
  license_link: LICENSE.md
5
  tags:
6
  - math
 
13
 
14
  # MathConstructOptimize-Envs (Harbor)
15
 
16
+ The 3,793 tasks of [`amphora/MathConstructOptimize-Envs`](https://huggingface.co/datasets/amphora/MathConstructOptimize-Envs) as runnable [Harbor](https://github.com/laude-institute/harbor) environments. Each task is one directory:
17
 
18
  ```
19
  tasks/<task_id>/
 
34
  ```
35
 
36
  **Validation:**
37
+ - The oracle agent passes on all 3,793 tasks (construct = 1.0, optimize ≥ 0.1), and a no-op agent scores 0.
38
  - An HTTP canary confirmed that `no-network` blocks internet access inside the task container. Research-level best constructions are public, so do not relax this for RL.
39
  - If you run many trials concurrently, raise Docker's default address pools. Each no-network task creates its own networks, and the defaults are exhausted at roughly 30 concurrent trials.
40
 
41
+ **v2 (2026-11-18):** 216 optimize tasks added (Optimization Constants in Mathematics, more AlphaEvolve problems, Leaps in Bounds, FrontierShift, FrontierCO MDS, ALE-Bench, FrontierCS, EdgeBench); the previous version is git tag `pre-v2`. ALE-Bench tasks are CC BY-ND 4.0 and FrontierShift tasks have no stated license; see LICENSE.md.
42
+
43
  See the main dataset card for the reward contract, tags, sources and the upstream verifier bugs we fixed.
families.csv CHANGED
@@ -1,143 +1,220 @@
1
- family,subset,tier,domain,source,license,problem_key,n_tasks,tags
2
- ae_p10_kakeya_3d,optimize,research,geometry,AlphaEvolve-Repository-of-Problems/10-3d-kakeya,CC-BY-4.0,kakeya_needle_3d_prisms,1,exact_piecewise_integration;float_objective
3
- ae_p16_gagliardo_nirenberg,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/16-gagliardo-nirenberg-special-case,CC-BY-4.0,nagy_gagliardo_nirenberg_sharp_constant,3,float_objective;known_optimum
4
- ae_p17_young_convolution,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/17-young-convolution,CC-BY-4.0,young_convolution_sharp_constant,3,float_objective;known_optimum
5
- ae_p18_hardy_littlewood_maximal,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/18-hardy-littlewood-maximal,CC-BY-4.0,centered_hardy_littlewood_weak11_constant_1d,1,exact_rational;known_optimum
6
- ae_p1_ff_kakeya,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/1-finite-field-kakeya,CC-BY-4.0,finite_field_kakeya_min_size,10,exact_integer;finite_fields
7
- ae_p1_ff_nikodym,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym,CC-BY-4.0,finite_field_nikodym_min_size_d3,4,exact_integer;finite_fields
8
- ae_p24_de_bruijn_sharma,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/24-de-bruijn-sharma,CC-BY-4.0,de_bruijn_sharma_x_intercept,4,float_objective;polynomials
9
- ae_p29_block_stacking,optimize,competition,analysis,AlphaEvolve-Repository-of-Problems/29-block-stacking,CC-BY-4.0,block_stacking_overhang_one_per_level,3,agentic_trivial;exact_rational;known_optimum
10
- ae_p30_arithmetic_kakeya,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/30-arithmetic-kakeya,CC-BY-4.0,arithmetic_kakeya_entropy_sum_difference,2,additive_combinatorics;entropy;float_objective
11
- ae_p31_furstenberg_sarkozy,optimize,research,number_theory,AlphaEvolve-Repository-of-Problems/31-furstenberg-sarkozy,CC-BY-4.0,furstenberg_sarkozy_modular_exponent,2,additive_combinatorics;exact_integer
12
- ae_p32_spherical_designs,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/32-spherical-designs,CC-BY-4.0,spherical_t_design_s2,6,float_tolerance;numerical_design
13
- ae_p35_cubes_in_cube,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/35-packing-in-a-dilate,CC-BY-4.0,unit_cubes_in_smallest_cube_n11,1,float_objective;packing
14
- ae_p37_turan_tetrahedron,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/37-hypergraph-turan-tetrahedron,CC-BY-4.0,turan_density_K4_3,1,blow_up;extremal_hypergraph
15
- ae_p38_factorial_factors,construct,research,number_theory,AlphaEvolve-Repository-of-Problems/38-factoring-N-factorial,CC-BY-4.0,factorial_into_n_large_factors_A034258,60,agentic_trivial;exact_integer;ilp_friendly;np_search
16
- ae_p39_beat_the_average,optimize,research,probability,AlphaEvolve-Repository-of-Problems/39-beat-the-average,CC-BY-4.0,beat_the_average_X1X2X3_lt_2X4,1,discrete_measure;float_objective
17
- ae_p40_erdos_discrepancy,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/40-erdos-discrepancy,CC-BY-4.0,erdos_discrepancy_C2_max_length,1,exact_integer;np_search;sat_friendly
18
- ae_p41_sphere_max_volume,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/41-points-on-sphere-max-volume,CC-BY-4.0,max_volume_inscribed_polyhedron_n_vertices,7,float_objective
19
- ae_p42_sum_difference_i,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/42-sum-difference-I,CC-BY-4.0,sum_difference_sumset_vs_diffset_exponent,1,additive_combinatorics;exact_integer
20
- ae_p43_sum_difference_ii,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/43-sum-difference-II,CC-BY-4.0,sum_difference_diffset_vs_sumset_exponent,1,additive_combinatorics;exact_integer
21
- ae_p44_sum_difference_iii,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/44-sum-difference-III,CC-BY-4.0,sum_difference_A_plus_B_vs_A_minus_B_exponent,1,additive_combinatorics;exact_integer
22
- ae_p45_sum_product_fp,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/45-sum-product-finite-field,Apache-2.0,finite_field_sum_product_sqrt_p,3,additive_combinatorics;exact_integer
23
- ae_p52_erdos_squarefree,construct,research,number_theory,AlphaEvolve-Repository-of-Problems/52-erdos-squarefree,CC-BY-4.0,erdos_ab_plus_1_never_squarefree,50,exact_integer
24
- ae_p54_touching_cylinders,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/54-pairwise-touching-cylinders,CC-BY-4.0,mutually_touching_infinite_cylinders,2,float_tolerance;nonlinear_system
25
- ae_p58_happy_ending,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/58-erdos-szekeres-happy-ending,CC-BY-4.0,erdos_szekeres_no_convex_k_gon_2pow_k_minus_2,5,exact_integer;np_search
26
- ae_p59_isosceles_free_grid,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/59-no-isosceles-grid,CC-BY-4.0,isosceles_free_grid_subset,2,exact_integer;np_search
27
- ae_p60_no5_on_sphere,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/60-no-5-on-a-sphere,CC-BY-4.0,no_five_cospherical_grid_subset,6,exact_integer;np_search
28
- ae_p65_imo2025_p6_tiling,construct,competition,combinatorics,AlphaEvolve-Repository-of-Problems/65-imo-2025-p6,CC-BY-4.0,imo2025_p6_min_tiles,6,exact_integer
29
- ea_circle_packing_26,optimize,research,discrete_geometry,EinsteinArena/circle-packing,MIT,circle_packing_square_sum_radii_n26,1,exact_check;float_objective
30
- ea_circles_rectangle_21,optimize,research,discrete_geometry,EinsteinArena/circles-rectangle,MIT,circles_in_rectangle_perimeter4_sum_radii_n21,1,exact_check;float_objective
31
- ea_difference_bases,optimize,research,number_theory,EinsteinArena/difference-bases,MIT,difference_basis_ratio,1,exact_check
32
- ea_edges_vs_triangles,optimize,research,combinatorics,EinsteinArena/edges-vs-triangles,MIT,edges_vs_triangles_density_curve,1,float_objective
33
- ea_erdos_min_overlap,optimize,research,analysis,EinsteinArena/erdos-min-overlap,MIT,erdos_minimum_overlap_upper,1,float_objective
34
- ea_first_autocorrelation,optimize,research,analysis,EinsteinArena/first-autocorrelation-inequality,MIT,first_autocorrelation_inequality_upper,1,float_objective
35
- ea_flat_polynomials_70,optimize,research,analysis,EinsteinArena/flat-polynomials,MIT,flat_littlewood_polynomial_n70,1,float_objective
36
- ea_hadamard_det_51,optimize,research,combinatorics,EinsteinArena/hadamard-det-51,MIT,hadamard_maxdet_n51,1,ea_under_review;exact_check
37
- ea_heilbronn_triangle_11,optimize,research,discrete_geometry,EinsteinArena/heilbronn-triangles,MIT,heilbronn_triangle_n11,1,exact_check;float_objective
38
- ea_kakeya_needle_128,optimize,research,geometry,EinsteinArena/kakeya-needle-128,MIT,kakeya_needle_triangles_n128,1,ea_under_review;exact_check
39
- ea_kissing_number,optimize,research,discrete_geometry,"EinsteinArena/kissing-number-d11,-d11-605,-d12",MIT,kissing_number_overlap_loss,3,best_known_uncertain;exact_check
40
- ea_min_distance_ratio_16,optimize,research,discrete_geometry,EinsteinArena/min-distance-ratio-2d,MIT,max_min_distance_ratio_2d_n16,1,float_objective
41
- ea_no_three_in_line_75,optimize,research,discrete_geometry,EinsteinArena/no-three-in-line-75,MIT,no_three_in_line_n75,1,ea_under_review;exact_check
42
- ea_prime_number_theorem,optimize,research,number_theory,EinsteinArena/prime-number-theorem,MIT,pnt_chebyshev_certificate,1,float_objective;range_limited_certificate
43
- ea_ring_loading_15,optimize,research,combinatorics,EinsteinArena/ring-loading-15,MIT,ring_loading_m15,1,ea_under_review;exact_check
44
- ea_second_autocorrelation,optimize,research,analysis,EinsteinArena/second-autocorrelation-inequality,MIT,second_autocorrelation_inequality_lower,1,float_objective
45
- ea_shannon_c7_5,optimize,research,graph_theory,EinsteinArena/shannon-capacity-c7-5,MIT,independent_set_c7_strong_power5,1,ea_under_review;exact_check
46
- ea_sidon_45_set,optimize,research,number_theory,EinsteinArena/sidon-45-set,MIT,sidon_subsets_of_45_sets,1,ea_under_review;exact_check
47
- ea_snake_in_the_box_13,optimize,research,graph_theory,EinsteinArena/snake-in-the-box-13,MIT,snake_in_the_box_d13,1,ea_under_review;exact_check
48
- ea_sorting_network_16,optimize,research,combinatorics,EinsteinArena/sorting-network-16,MIT,sorting_network_size_n16,1,ea_under_review;exact_check
49
- ea_spencer_discrepancy,optimize,research,combinatorics,EinsteinArena/spencer-discrepancy,MIT,spencer_discrepancy_square_n_le20,1,ea_under_review;exact_check
50
- ea_tammes,optimize,research,discrete_geometry,EinsteinArena/tammes-problem,MIT,tammes_n50,1,float_objective
51
- ea_third_autocorrelation,optimize,research,analysis,EinsteinArena/third-autocorrelation-inequality,MIT,third_autocorrelation_inequality_upper,1,float_objective
52
- ea_thomson_282,optimize,research,discrete_geometry,EinsteinArena/thomson-problem,MIT,thomson_problem_n282,1,float_objective
53
- ea_two_deletion_code_16,optimize,research,coding_theory,EinsteinArena/two-deletion-code-16,MIT,two_deletion_code_n16,1,ea_under_review;exact_check
54
- ea_uncertainty_principle,optimize,research,analysis,EinsteinArena/uncertainty-principle,MIT,plus1_uncertainty_principle_upper,1,high_precision
55
- finch_3ap_diffs_40,optimize,research,combinatorics,Finch-EFT/skydiscover/erdos_1097_3ap_diffs_n40,Apache-2.0,erdos1097_3ap_differences_n40_m400,1,best_known_uncertain;exact_check;np_search
56
- finch_4subset_3dist_16,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_659_4subset_3dist_construction,Apache-2.0,erdos659_4sets_3dist_n16,1,best_known_uncertain;exact_check;np_search
57
- finch_flat_poly_32,optimize,research,analysis,Finch-EFT/skydiscover/erdos_flat_polynomials_n32,Apache-2.0,flat_littlewood_polynomial_n32,1,best_known_uncertain;float_objective;np_search
58
- finch_heilbronn_disk_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_507_heilbronn_disk_n12,Apache-2.0,heilbronn_disk_n12,1,best_known_uncertain;exact_check;float_objective
59
- finch_isosceles_free_grid_8,optimize,competition,discrete_geometry,Finch-EFT/skydiscover/erdos_isosceles_free_grid,Apache-2.0,isosceles_free_grid_subset,1,exact_check;exact_optimum_known;np_search
60
- finch_lcm_chain_10000,optimize,competition,number_theory,Finch-EFT/skydiscover/erdos_440_lcm_density,Apache-2.0,erdos440_lcm_consecutive_x10000,1,exact_check;exact_optimum_known
61
- finch_newman_200,optimize,research,number_theory,Finch-EFT/skydiscover/erdos_480_newman_density,Apache-2.0,erdos480_newman_finite_n200,1,best_known_uncertain;float_objective
62
- finch_no4_concyclic_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_654_no_four_concyclic_distances,Apache-2.0,erdos654_no4_concyclic_n12,1,best_known_uncertain;exact_check;np_search
63
- finch_sidon_pair_100,optimize,research,number_theory,Finch-EFT/skydiscover/erdos_43_sidon_pair,Apache-2.0,erdos43_sidon_pair_n100,1,best_known_uncertain;exact_check;np_search
64
- finch_sidon_set_121,optimize,competition,number_theory,Finch-EFT/skydiscover/erdos_difference_bases_N121,Apache-2.0,max_sidon_set_0_120,1,exact_check;exact_optimum_known
65
- finch_square_packing_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_106_square_packing,Apache-2.0,erdos106_squares_in_square_n12,1,exact_check
66
- finch_udg_girth_30,optimize,research,graph_theory,Finch-EFT/skydiscover/erdos_705_unit_distance_high_girth,Apache-2.0,erdos705_udg_chi4_girth_n30,1,best_known_uncertain;exact_check
67
- fs_admissible_set,optimize,research,combinatorics,FunSearch/admissible_set,Apache-2.0,admissible_set_constant_weight,2,integer_objective
68
- fs_cap_set,optimize,research,combinatorics,FunSearch/cap_set,Apache-2.0,cap_set_f3n,6,integer_objective
69
- fs_cycle_independent_set,optimize,research,graph_theory,FunSearch/cyclic_graphs,Apache-2.0,independent_set_strong_power_odd_cycle,7,integer_objective
70
- mc_all_interval,construct,competition,combinatorics,MathConstraint/all_interval,CC-BY-4.0,all_interval_series,40,agentic_trivial
71
- mc_antimagic_square,construct,competition,combinatorics,MathConstraint/antimagic_square,CC-BY-4.0,antimagic_square,6,np_search
72
- mc_bibd,construct,competition,combinatorics,MathConstraint/bibd,CC-BY-4.0,bibd,28,np_search
73
- mc_costas_array,construct,competition,combinatorics,MathConstraint/costas_array,CC-BY-4.0,costas_array,23,np_search
74
- mc_debruijn,construct,competition,combinatorics,MathConstraint/debruijn,CC-BY-4.0,de_bruijn_sequence,22,agentic_trivial
75
- mc_golomb,construct,competition,combinatorics,MathConstraint/golomb,CC-BY-4.0,golomb_ruler,18,np_search
76
- mc_graceful_graph,construct,competition,graph_theory,MathConstraint/graceful_graph,CC-BY-4.0,graceful_labeling,17,np_search
77
- mc_graph_k_coloring,construct,competition,graph_theory,MathConstraint/graph_k_coloring,CC-BY-4.0,graph_k_coloring,48,np_search
78
- mc_hadamard,construct,competition,combinatorics,MathConstraint/hadamard,CC-BY-4.0,legendre_pair,30,np_search
79
- mc_langford,construct,competition,combinatorics,MathConstraint/langford,CC-BY-4.0,langford_pairing,27,agentic_trivial;np_search
80
- mc_low_autocorrelation,construct,competition,combinatorics,MathConstraint/low_autocorrelation,CC-BY-4.0,low_autocorrelation_binary_sequence,31,np_search
81
- mc_magic_sequence,construct,competition,combinatorics,MathConstraint/magic_sequence,CC-BY-4.0,magic_sequence,30,agentic_trivial
82
- mc_max_clique,construct,competition,graph_theory,MathConstraint/max_clique,CC-BY-4.0,graph_clique,55,np_search
83
- mc_max_independent_set,construct,competition,graph_theory,MathConstraint/max_independent_set,CC-BY-4.0,graph_independent_set,60,np_search
84
- mc_non_transitive_dice,construct,competition,probability,MathConstraint/non_transitive_dice,CC-BY-4.0,intransitive_dice,15,agentic_trivial;np_search
85
- mc_number_partitioning,construct,competition,number_theory,MathConstraint/number_partitioning,CC-BY-4.0,equal_sum_partition_1_to_n,59,agentic_trivial
86
- mc_ortholatin,construct,competition,combinatorics,MathConstraint/ortholatin,CC-BY-4.0,orthogonal_latin_squares,21,np_search
87
- mc_pysms_chromatic_girth,construct,competition,graph_theory,MathConstraint/pysms_chromatic_girth,CC-BY-4.0,graph_existence_chromatic_girth,25,np_search
88
- mc_pysms_clique_coloring,construct,competition,graph_theory,MathConstraint/pysms_clique_coloring,CC-BY-4.0,graph_existence_clique_coloring,16,np_search
89
- mc_pysms_degree_bounds,construct,competition,graph_theory,MathConstraint/pysms_degree_bounds,CC-BY-4.0,graph_existence_degree_bounds,11,agentic_trivial;np_search
90
- mc_pysms_girth_degree,construct,competition,graph_theory,MathConstraint/pysms_girth_degree,CC-BY-4.0,graph_existence_girth_degree,18,np_search
91
- mc_pysms_graph_builder,construct,competition,graph_theory,MathConstraint/pysms_graph_builder,CC-BY-4.0,graph_existence_graph_builder,8,np_search
92
- mc_pysms_independent_connectivity,construct,competition,graph_theory,MathConstraint/pysms_independent_connectivity,CC-BY-4.0,graph_existence_independent_connectivity,15,np_search
93
- mc_pysms_min_connectivity,construct,competition,graph_theory,MathConstraint/pysms_min_connectivity,CC-BY-4.0,graph_existence_min_connectivity,12,agentic_trivial;np_search
94
- mc_pysms_min_degree,construct,competition,graph_theory,MathConstraint/pysms_min_degree,CC-BY-4.0,graph_existence_min_degree,12,agentic_trivial;np_search
95
- mc_pysms_min_girth,construct,competition,graph_theory,MathConstraint/pysms_min_girth,CC-BY-4.0,graph_existence_min_girth,13,agentic_trivial;np_search
96
- mc_pysms_mtf,construct,competition,graph_theory,MathConstraint/pysms_mtf,CC-BY-4.0,maximal_triangle_free_graph,12,agentic_trivial;np_search
97
- mc_pysms_num_edges_bounds,construct,competition,graph_theory,MathConstraint/pysms_num_edges_bounds,CC-BY-4.0,graph_existence_num_edges_bounds,12,agentic_trivial;np_search
98
- mc_quasigroup_idempotent,construct,competition,algebra,MathConstraint/quasigroup_idempotent,CC-BY-4.0,idempotent_qg3_quasigroup,12,np_search
99
- mc_ramsey,construct,competition,combinatorics,MathConstraint/ramsey,CC-BY-4.0,ramsey_two_colouring,54,np_search
100
- mc_social_golfers,construct,competition,combinatorics,MathConstraint/social_golfers,CC-BY-4.0,social_golfer,23,np_search
101
- mc_van_der_waerden,construct,competition,combinatorics,MathConstraint/van_der_waerden,CC-BY-4.0,van_der_waerden_colouring,42,np_search
102
- mc_vertex_cover,construct,competition,graph_theory,MathConstraint/vertex_cover,CC-BY-4.0,graph_vertex_cover,60,np_search
103
- nppc_bandwidth,construct,competition,graph_theory,NPPC/bandwidth,MIT,graph_bandwidth,60,agentic_trivial;np_search
104
- nppc_clique,construct,competition,graph_theory,NPPC/clique,MIT,graph_clique,60,agentic_trivial;np_search
105
- nppc_dominating_set,construct,competition,graph_theory,NPPC/dominating_set,MIT,graph_dominating_set,60,agentic_trivial;np_search
106
- nppc_independent_set,construct,competition,graph_theory,NPPC/independent_set,MIT,graph_independent_set,60,agentic_trivial;np_search
107
- nppc_quad_diophantine,construct,competition,number_theory,NPPC/quad_diop_equ,MIT,quadratic_diophantine_ax2_by_c,60,agentic_trivial;np_search
108
- nppc_quadratic_congruence,construct,competition,number_theory,NPPC/quadratic_congruence,MIT,quadratic_congruence_bounded,60,agentic_trivial;np_search
109
- nppc_three_coloring,construct,competition,graph_theory,NPPC/graph_three_colorability,MIT,graph_3_coloring,60,agentic_trivial;np_search
110
- nppc_vertex_cover,construct,competition,graph_theory,NPPC/vertex_cover,MIT,graph_vertex_cover,60,agentic_trivial;np_search
111
- rg_graph_color,construct,competition,graph_theory,ReasoningGym/graph_color,Apache-2.0,graph_k_coloring,60,agentic_trivial;np_search
112
- rg_intermediate_integration,construct,competition,analysis,ReasoningGym/intermediate_integration,Apache-2.0,symbolic_antiderivative,60,symbolic
113
- rg_simple_integration,construct,competition,analysis,ReasoningGym/simple_integration,Apache-2.0,symbolic_antiderivative_polynomial,60,symbolic
114
- rlve_antiderivative,construct,competition,analysis,RLVE-Gym/integral,MIT,symbolic_antiderivative,60,
115
- rlve_bezout_identity,construct,competition,number_theory,RLVE-Gym/bezout_identity,MIT,bezout_coefficients,60,agentic_trivial
116
- rlve_construct_hack_interval,construct,competition,number_theory,RLVE-Gym/construct_hack_interval,MIT,cf468c_hack_interval,60,
117
- rlve_crt,construct,competition,number_theory,RLVE-Gym/crt,MIT,generalized_crt,60,agentic_trivial
118
- rlve_distinct_subset_sum_permutation,construct,competition,combinatorics,RLVE-Gym/distinct_array_permutation,MIT,cf892d_gluttony,60,agentic_trivial
119
- rlve_even_degree_partition,construct,competition,graph_theory,RLVE-Gym/even_degree_graph_partitioning,MIT,gallai_even_partition,60,agentic_trivial
120
- rlve_graph_isomorphism,construct,competition,graph_theory,RLVE-Gym/graph_isomorphism,MIT,graph_isomorphism,60,agentic_trivial;np_search
121
- rlve_grid_parity_construction,construct,competition,algebra,RLVE-Gym/grid_parity_construction,MIT,lights_out_gf2,60,agentic_trivial
122
- rlve_imp_party,construct,competition,graph_theory,RLVE-Gym/imp_party,MIT,poi_party_clique_3n,60,agentic_trivial
123
- rlve_integer_linear_system,construct,competition,algebra,RLVE-Gym/gaussian_elimination,MIT,integer_linear_system,60,agentic_trivial
124
- rlve_largest_convex_polygon,construct,competition,discrete_geometry,RLVE-Gym/largest_convex_polygon,MIT,largest_convex_subset,60,agentic_trivial
125
- rlve_largest_rectangle,construct,competition,discrete_geometry,RLVE-Gym/largest_rectangle_among_points,MIT,largest_rectangle_in_point_set,60,agentic_trivial
126
- rlve_matrix_permutation_equivalence,construct,competition,combinatorics,RLVE-Gym/matrix_permutation_equivalence,MIT,binary_matrix_permutation_equivalence,60,agentic_trivial;np_search
127
- rlve_max_achromatic_coloring,construct,competition,graph_theory,RLVE-Gym/maximum_achromatic_number,MIT,achromatic_number,60,agentic_trivial;np_search
128
- rlve_maximum_clique,construct,competition,graph_theory,RLVE-Gym/maximum_clique,MIT,maximum_clique,60,agentic_trivial;np_search
129
- rlve_min_chromatic_number,construct,competition,graph_theory,RLVE-Gym/minimum_chromatic_number,MIT,graph_k_coloring_chromatic,60,agentic_trivial;np_search
130
- rlve_min_harmonious_coloring,construct,competition,graph_theory,RLVE-Gym/minimum_harmonious_chromatic_number,MIT,harmonious_chromatic_number,60,agentic_trivial;np_search
131
- rlve_pairwise_sums_recovery,construct,competition,combinatorics,RLVE-Gym/squ_squarks,MIT,poi_squarks_pairwise_sums,60,agentic_trivial
132
- rlve_pan_solar_panels,construct,competition,number_theory,RLVE-Gym/pan_solar_panels,MIT,poi_solar_panels_max_gcd,60,agentic_trivial
133
- rlve_polynomial_global_minimum,construct,competition,analysis,RLVE-Gym/polynomial_minimum,MIT,polynomial_global_minimum,60,agentic_trivial
134
- rlve_polynomial_integer_roots,construct,competition,algebra,RLVE-Gym/polynomial_factorization,MIT,integer_root_factorization,60,agentic_trivial;unique_answer
135
- rlve_polynomial_interpolation,construct,competition,algebra,RLVE-Gym/polynomial_interpolation,MIT,lagrange_interpolation_integer,60,agentic_trivial;unique_answer
136
- rlve_prefix_product_permutation,construct,competition,number_theory,RLVE-Gym/prefix_product_mod_distinct_permutation,MIT,cf487c_prefix_product_sequence,60,
137
- rlve_prefix_sum_permutation,construct,competition,number_theory,RLVE-Gym/prefix_sum_mod_distinct_permutation,MIT,prefix_sum_mod_distinct_permutation,60,
138
- rlve_prime_graph_coloring,construct,competition,graph_theory,RLVE-Gym/prime_graph_minimum_chromatic_number,MIT,prime_distance_graph_coloring,60,agentic_trivial
139
- rlve_recursive_sequence_sum,construct,competition,number_theory,RLVE-Gym/recursive_sequence_sum_construction,MIT,linear_recurrence_subset_sum,60,agentic_trivial
140
- rlve_round_robin_scores,construct,competition,combinatorics,RLVE-Gym/round_robin,MIT,tournament_score_realisation_3_1_0,60,agentic_trivial
141
- rlve_shortest_path_count,construct,competition,combinatorics,RLVE-Gym/shortest_path_count_construction,MIT,cf388b_shortest_path_count,60,
142
- rlve_smallest_enclosing_circle,construct,competition,geometry,RLVE-Gym/smallest_circle,MIT,minimum_enclosing_circle,60,agentic_trivial
143
- rlve_subgraph_isomorphism,construct,competition,graph_theory,RLVE-Gym/subgraph_isomorphism,MIT,induced_subgraph_isomorphism,60,agentic_trivial;np_search
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ family,subset,tier,domain,source,license,problem_key,n_tasks,tags
2
+ ae_p10_kakeya_3d,optimize,research,geometry,AlphaEvolve-Repository-of-Problems/10-3d-kakeya,CC-BY-4.0,kakeya_needle_3d_prisms,1,exact_piecewise_integration;float_objective
3
+ ae_p16_gagliardo_nirenberg,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/16-gagliardo-nirenberg-special-case,CC-BY-4.0,nagy_gagliardo_nirenberg_sharp_constant,3,float_objective;known_optimum
4
+ ae_p17_young_convolution,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/17-young-convolution,CC-BY-4.0,young_convolution_sharp_constant,3,float_objective;known_optimum
5
+ ae_p18_hardy_littlewood_maximal,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/18-hardy-littlewood-maximal,CC-BY-4.0,centered_hardy_littlewood_weak11_constant_1d,1,exact_rational;known_optimum
6
+ ae_p1_ff_kakeya,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/1-finite-field-kakeya,CC-BY-4.0,finite_field_kakeya_min_size,10,exact_integer;finite_fields
7
+ ae_p1_ff_nikodym,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym,CC-BY-4.0,finite_field_nikodym_min_size_d3,4,exact_integer;finite_fields
8
+ ae_p24_de_bruijn_sharma,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/24-de-bruijn-sharma,CC-BY-4.0,de_bruijn_sharma_x_intercept,4,float_objective;polynomials
9
+ ae_p29_block_stacking,optimize,competition,analysis,AlphaEvolve-Repository-of-Problems/29-block-stacking,CC-BY-4.0,block_stacking_overhang_one_per_level,3,agentic_trivial;exact_rational;known_optimum
10
+ ae_p30_arithmetic_kakeya,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/30-arithmetic-kakeya,CC-BY-4.0,arithmetic_kakeya_entropy_sum_difference,2,additive_combinatorics;entropy;float_objective
11
+ ae_p31_furstenberg_sarkozy,optimize,research,number_theory,AlphaEvolve-Repository-of-Problems/31-furstenberg-sarkozy,CC-BY-4.0,furstenberg_sarkozy_modular_exponent,2,additive_combinatorics;exact_integer
12
+ ae_p32_spherical_designs,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/32-spherical-designs,CC-BY-4.0,spherical_t_design_s2,6,float_tolerance;numerical_design
13
+ ae_p35_cubes_in_cube,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/35-packing-in-a-dilate,CC-BY-4.0,unit_cubes_in_smallest_cube_n11,1,float_objective;packing
14
+ ae_p37_turan_tetrahedron,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/37-hypergraph-turan-tetrahedron,CC-BY-4.0,turan_density_K4_3,1,blow_up;extremal_hypergraph
15
+ ae_p38_factorial_factors,construct,research,number_theory,AlphaEvolve-Repository-of-Problems/38-factoring-N-factorial,CC-BY-4.0,factorial_into_n_large_factors_A034258,60,agentic_trivial;exact_integer;ilp_friendly;np_search
16
+ ae_p39_beat_the_average,optimize,research,probability,AlphaEvolve-Repository-of-Problems/39-beat-the-average,CC-BY-4.0,beat_the_average_X1X2X3_lt_2X4,1,discrete_measure;float_objective
17
+ ae_p40_erdos_discrepancy,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/40-erdos-discrepancy,CC-BY-4.0,erdos_discrepancy_C2_max_length,1,exact_integer;np_search;sat_friendly
18
+ ae_p41_sphere_max_volume,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/41-points-on-sphere-max-volume,CC-BY-4.0,max_volume_inscribed_polyhedron_n_vertices,7,float_objective
19
+ ae_p42_sum_difference_i,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/42-sum-difference-I,CC-BY-4.0,sum_difference_sumset_vs_diffset_exponent,1,additive_combinatorics;exact_integer
20
+ ae_p43_sum_difference_ii,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/43-sum-difference-II,CC-BY-4.0,sum_difference_diffset_vs_sumset_exponent,1,additive_combinatorics;exact_integer
21
+ ae_p44_sum_difference_iii,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/44-sum-difference-III,CC-BY-4.0,sum_difference_A_plus_B_vs_A_minus_B_exponent,1,additive_combinatorics;exact_integer
22
+ ae_p45_sum_product_fp,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/45-sum-product-finite-field,Apache-2.0,finite_field_sum_product_sqrt_p,3,additive_combinatorics;exact_integer
23
+ ae_p52_erdos_squarefree,construct,research,number_theory,AlphaEvolve-Repository-of-Problems/52-erdos-squarefree,CC-BY-4.0,erdos_ab_plus_1_never_squarefree,50,exact_integer
24
+ ae_p54_touching_cylinders,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/54-pairwise-touching-cylinders,CC-BY-4.0,mutually_touching_infinite_cylinders,2,float_tolerance;nonlinear_system
25
+ ae_p58_happy_ending,construct,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/58-erdos-szekeres-happy-ending,CC-BY-4.0,erdos_szekeres_no_convex_k_gon_2pow_k_minus_2,5,exact_integer;np_search
26
+ ae_p59_isosceles_free_grid,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/59-no-isosceles-grid,CC-BY-4.0,isosceles_free_grid_subset,2,exact_integer;np_search
27
+ ae_p60_no5_on_sphere,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/60-no-5-on-a-sphere,CC-BY-4.0,no_five_cospherical_grid_subset,6,exact_integer;np_search
28
+ ae_p65_imo2025_p6_tiling,construct,competition,combinatorics,AlphaEvolve-Repository-of-Problems/65-imo-2025-p6,CC-BY-4.0,imo2025_p6_min_tiles,6,exact_integer
29
+ ea_circle_packing_26,optimize,research,discrete_geometry,EinsteinArena/circle-packing,MIT,circle_packing_square_sum_radii_n26,1,exact_check;float_objective
30
+ ea_circles_rectangle_21,optimize,research,discrete_geometry,EinsteinArena/circles-rectangle,MIT,circles_in_rectangle_perimeter4_sum_radii_n21,1,exact_check;float_objective
31
+ ea_difference_bases,optimize,research,number_theory,EinsteinArena/difference-bases,MIT,difference_basis_ratio,1,exact_check
32
+ ea_edges_vs_triangles,optimize,research,combinatorics,EinsteinArena/edges-vs-triangles,MIT,edges_vs_triangles_density_curve,1,float_objective
33
+ ea_erdos_min_overlap,optimize,research,analysis,EinsteinArena/erdos-min-overlap,MIT,erdos_minimum_overlap_upper,1,float_objective
34
+ ea_first_autocorrelation,optimize,research,analysis,EinsteinArena/first-autocorrelation-inequality,MIT,first_autocorrelation_inequality_upper,1,float_objective
35
+ ea_flat_polynomials_70,optimize,research,analysis,EinsteinArena/flat-polynomials,MIT,flat_littlewood_polynomial_n70,1,float_objective
36
+ ea_hadamard_det_51,optimize,research,combinatorics,EinsteinArena/hadamard-det-51,MIT,hadamard_maxdet_n51,1,ea_under_review;exact_check
37
+ ea_heilbronn_triangle_11,optimize,research,discrete_geometry,EinsteinArena/heilbronn-triangles,MIT,heilbronn_triangle_n11,1,exact_check;float_objective
38
+ ea_kakeya_needle_128,optimize,research,geometry,EinsteinArena/kakeya-needle-128,MIT,kakeya_needle_triangles_n128,1,ea_under_review;exact_check
39
+ ea_kissing_number,optimize,research,discrete_geometry,"EinsteinArena/kissing-number-d11,-d11-605,-d12",MIT,kissing_number_overlap_loss,3,best_known_uncertain;exact_check
40
+ ea_min_distance_ratio_16,optimize,research,discrete_geometry,EinsteinArena/min-distance-ratio-2d,MIT,max_min_distance_ratio_2d_n16,1,float_objective
41
+ ea_no_three_in_line_75,optimize,research,discrete_geometry,EinsteinArena/no-three-in-line-75,MIT,no_three_in_line_n75,1,ea_under_review;exact_check
42
+ ea_prime_number_theorem,optimize,research,number_theory,EinsteinArena/prime-number-theorem,MIT,pnt_chebyshev_certificate,1,float_objective;range_limited_certificate
43
+ ea_ring_loading_15,optimize,research,combinatorics,EinsteinArena/ring-loading-15,MIT,ring_loading_m15,1,ea_under_review;exact_check
44
+ ea_second_autocorrelation,optimize,research,analysis,EinsteinArena/second-autocorrelation-inequality,MIT,second_autocorrelation_inequality_lower,1,float_objective
45
+ ea_shannon_c7_5,optimize,research,graph_theory,EinsteinArena/shannon-capacity-c7-5,MIT,independent_set_c7_strong_power5,1,ea_under_review;exact_check
46
+ ea_sidon_45_set,optimize,research,number_theory,EinsteinArena/sidon-45-set,MIT,sidon_subsets_of_45_sets,1,ea_under_review;exact_check
47
+ ea_snake_in_the_box_13,optimize,research,graph_theory,EinsteinArena/snake-in-the-box-13,MIT,snake_in_the_box_d13,1,ea_under_review;exact_check
48
+ ea_sorting_network_16,optimize,research,combinatorics,EinsteinArena/sorting-network-16,MIT,sorting_network_size_n16,1,ea_under_review;exact_check
49
+ ea_spencer_discrepancy,optimize,research,combinatorics,EinsteinArena/spencer-discrepancy,MIT,spencer_discrepancy_square_n_le20,1,ea_under_review;exact_check
50
+ ea_tammes,optimize,research,discrete_geometry,EinsteinArena/tammes-problem,MIT,tammes_n50,1,float_objective
51
+ ea_third_autocorrelation,optimize,research,analysis,EinsteinArena/third-autocorrelation-inequality,MIT,third_autocorrelation_inequality_upper,1,float_objective
52
+ ea_thomson_282,optimize,research,discrete_geometry,EinsteinArena/thomson-problem,MIT,thomson_problem_n282,1,float_objective
53
+ ea_two_deletion_code_16,optimize,research,coding_theory,EinsteinArena/two-deletion-code-16,MIT,two_deletion_code_n16,1,ea_under_review;exact_check
54
+ ea_uncertainty_principle,optimize,research,analysis,EinsteinArena/uncertainty-principle,MIT,plus1_uncertainty_principle_upper,1,high_precision
55
+ finch_3ap_diffs_40,optimize,research,combinatorics,Finch-EFT/skydiscover/erdos_1097_3ap_diffs_n40,Apache-2.0,erdos1097_3ap_differences_n40_m400,1,best_known_uncertain;exact_check;np_search
56
+ finch_4subset_3dist_16,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_659_4subset_3dist_construction,Apache-2.0,erdos659_4sets_3dist_n16,1,best_known_uncertain;exact_check;np_search
57
+ finch_flat_poly_32,optimize,research,analysis,Finch-EFT/skydiscover/erdos_flat_polynomials_n32,Apache-2.0,flat_littlewood_polynomial_n32,1,best_known_uncertain;float_objective;np_search
58
+ finch_heilbronn_disk_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_507_heilbronn_disk_n12,Apache-2.0,heilbronn_disk_n12,1,best_known_uncertain;exact_check;float_objective
59
+ finch_isosceles_free_grid_8,optimize,competition,discrete_geometry,Finch-EFT/skydiscover/erdos_isosceles_free_grid,Apache-2.0,isosceles_free_grid_subset,1,exact_check;exact_optimum_known;np_search
60
+ finch_lcm_chain_10000,optimize,competition,number_theory,Finch-EFT/skydiscover/erdos_440_lcm_density,Apache-2.0,erdos440_lcm_consecutive_x10000,1,exact_check;exact_optimum_known
61
+ finch_newman_200,optimize,research,number_theory,Finch-EFT/skydiscover/erdos_480_newman_density,Apache-2.0,erdos480_newman_finite_n200,1,best_known_uncertain;float_objective
62
+ finch_no4_concyclic_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_654_no_four_concyclic_distances,Apache-2.0,erdos654_no4_concyclic_n12,1,best_known_uncertain;exact_check;np_search
63
+ finch_sidon_pair_100,optimize,research,number_theory,Finch-EFT/skydiscover/erdos_43_sidon_pair,Apache-2.0,erdos43_sidon_pair_n100,1,best_known_uncertain;exact_check;np_search
64
+ finch_sidon_set_121,optimize,competition,number_theory,Finch-EFT/skydiscover/erdos_difference_bases_N121,Apache-2.0,max_sidon_set_0_120,1,exact_check;exact_optimum_known
65
+ finch_square_packing_12,optimize,research,discrete_geometry,Finch-EFT/skydiscover/erdos_106_square_packing,Apache-2.0,erdos106_squares_in_square_n12,1,exact_check
66
+ finch_udg_girth_30,optimize,research,graph_theory,Finch-EFT/skydiscover/erdos_705_unit_distance_high_girth,Apache-2.0,erdos705_udg_chi4_girth_n30,1,best_known_uncertain;exact_check
67
+ fs_admissible_set,optimize,research,combinatorics,FunSearch/admissible_set,Apache-2.0,admissible_set_constant_weight,2,integer_objective
68
+ fs_cap_set,optimize,research,combinatorics,FunSearch/cap_set,Apache-2.0,cap_set_f3n,6,integer_objective
69
+ fs_cycle_independent_set,optimize,research,graph_theory,FunSearch/cyclic_graphs,Apache-2.0,independent_set_strong_power_odd_cycle,7,integer_objective
70
+ mc_all_interval,construct,competition,combinatorics,MathConstraint/all_interval,CC-BY-4.0,all_interval_series,40,agentic_trivial
71
+ mc_antimagic_square,construct,competition,combinatorics,MathConstraint/antimagic_square,CC-BY-4.0,antimagic_square,6,np_search
72
+ mc_bibd,construct,competition,combinatorics,MathConstraint/bibd,CC-BY-4.0,bibd,28,np_search
73
+ mc_costas_array,construct,competition,combinatorics,MathConstraint/costas_array,CC-BY-4.0,costas_array,23,np_search
74
+ mc_debruijn,construct,competition,combinatorics,MathConstraint/debruijn,CC-BY-4.0,de_bruijn_sequence,22,agentic_trivial
75
+ mc_golomb,construct,competition,combinatorics,MathConstraint/golomb,CC-BY-4.0,golomb_ruler,18,np_search
76
+ mc_graceful_graph,construct,competition,graph_theory,MathConstraint/graceful_graph,CC-BY-4.0,graceful_labeling,17,np_search
77
+ mc_graph_k_coloring,construct,competition,graph_theory,MathConstraint/graph_k_coloring,CC-BY-4.0,graph_k_coloring,48,np_search
78
+ mc_hadamard,construct,competition,combinatorics,MathConstraint/hadamard,CC-BY-4.0,legendre_pair,30,np_search
79
+ mc_langford,construct,competition,combinatorics,MathConstraint/langford,CC-BY-4.0,langford_pairing,27,agentic_trivial;np_search
80
+ mc_low_autocorrelation,construct,competition,combinatorics,MathConstraint/low_autocorrelation,CC-BY-4.0,low_autocorrelation_binary_sequence,31,np_search
81
+ mc_magic_sequence,construct,competition,combinatorics,MathConstraint/magic_sequence,CC-BY-4.0,magic_sequence,30,agentic_trivial
82
+ mc_max_clique,construct,competition,graph_theory,MathConstraint/max_clique,CC-BY-4.0,graph_clique,55,np_search
83
+ mc_max_independent_set,construct,competition,graph_theory,MathConstraint/max_independent_set,CC-BY-4.0,graph_independent_set,60,np_search
84
+ mc_non_transitive_dice,construct,competition,probability,MathConstraint/non_transitive_dice,CC-BY-4.0,intransitive_dice,15,agentic_trivial;np_search
85
+ mc_number_partitioning,construct,competition,number_theory,MathConstraint/number_partitioning,CC-BY-4.0,equal_sum_partition_1_to_n,59,agentic_trivial
86
+ mc_ortholatin,construct,competition,combinatorics,MathConstraint/ortholatin,CC-BY-4.0,orthogonal_latin_squares,21,np_search
87
+ mc_pysms_chromatic_girth,construct,competition,graph_theory,MathConstraint/pysms_chromatic_girth,CC-BY-4.0,graph_existence_chromatic_girth,25,np_search
88
+ mc_pysms_clique_coloring,construct,competition,graph_theory,MathConstraint/pysms_clique_coloring,CC-BY-4.0,graph_existence_clique_coloring,16,np_search
89
+ mc_pysms_degree_bounds,construct,competition,graph_theory,MathConstraint/pysms_degree_bounds,CC-BY-4.0,graph_existence_degree_bounds,11,agentic_trivial;np_search
90
+ mc_pysms_girth_degree,construct,competition,graph_theory,MathConstraint/pysms_girth_degree,CC-BY-4.0,graph_existence_girth_degree,18,np_search
91
+ mc_pysms_graph_builder,construct,competition,graph_theory,MathConstraint/pysms_graph_builder,CC-BY-4.0,graph_existence_graph_builder,8,np_search
92
+ mc_pysms_independent_connectivity,construct,competition,graph_theory,MathConstraint/pysms_independent_connectivity,CC-BY-4.0,graph_existence_independent_connectivity,15,np_search
93
+ mc_pysms_min_connectivity,construct,competition,graph_theory,MathConstraint/pysms_min_connectivity,CC-BY-4.0,graph_existence_min_connectivity,12,agentic_trivial;np_search
94
+ mc_pysms_min_degree,construct,competition,graph_theory,MathConstraint/pysms_min_degree,CC-BY-4.0,graph_existence_min_degree,12,agentic_trivial;np_search
95
+ mc_pysms_min_girth,construct,competition,graph_theory,MathConstraint/pysms_min_girth,CC-BY-4.0,graph_existence_min_girth,13,agentic_trivial;np_search
96
+ mc_pysms_mtf,construct,competition,graph_theory,MathConstraint/pysms_mtf,CC-BY-4.0,maximal_triangle_free_graph,12,agentic_trivial;np_search
97
+ mc_pysms_num_edges_bounds,construct,competition,graph_theory,MathConstraint/pysms_num_edges_bounds,CC-BY-4.0,graph_existence_num_edges_bounds,12,agentic_trivial;np_search
98
+ mc_quasigroup_idempotent,construct,competition,algebra,MathConstraint/quasigroup_idempotent,CC-BY-4.0,idempotent_qg3_quasigroup,12,np_search
99
+ mc_ramsey,construct,competition,combinatorics,MathConstraint/ramsey,CC-BY-4.0,ramsey_two_colouring,54,np_search
100
+ mc_social_golfers,construct,competition,combinatorics,MathConstraint/social_golfers,CC-BY-4.0,social_golfer,23,np_search
101
+ mc_van_der_waerden,construct,competition,combinatorics,MathConstraint/van_der_waerden,CC-BY-4.0,van_der_waerden_colouring,42,np_search
102
+ mc_vertex_cover,construct,competition,graph_theory,MathConstraint/vertex_cover,CC-BY-4.0,graph_vertex_cover,60,np_search
103
+ nppc_bandwidth,construct,competition,graph_theory,NPPC/bandwidth,MIT,graph_bandwidth,60,agentic_trivial;np_search
104
+ nppc_clique,construct,competition,graph_theory,NPPC/clique,MIT,graph_clique,60,agentic_trivial;np_search
105
+ nppc_dominating_set,construct,competition,graph_theory,NPPC/dominating_set,MIT,graph_dominating_set,60,agentic_trivial;np_search
106
+ nppc_independent_set,construct,competition,graph_theory,NPPC/independent_set,MIT,graph_independent_set,60,agentic_trivial;np_search
107
+ nppc_quad_diophantine,construct,competition,number_theory,NPPC/quad_diop_equ,MIT,quadratic_diophantine_ax2_by_c,60,agentic_trivial;np_search
108
+ nppc_quadratic_congruence,construct,competition,number_theory,NPPC/quadratic_congruence,MIT,quadratic_congruence_bounded,60,agentic_trivial;np_search
109
+ nppc_three_coloring,construct,competition,graph_theory,NPPC/graph_three_colorability,MIT,graph_3_coloring,60,agentic_trivial;np_search
110
+ nppc_vertex_cover,construct,competition,graph_theory,NPPC/vertex_cover,MIT,graph_vertex_cover,60,agentic_trivial;np_search
111
+ rg_graph_color,construct,competition,graph_theory,ReasoningGym/graph_color,Apache-2.0,graph_k_coloring,60,agentic_trivial;np_search
112
+ rg_intermediate_integration,construct,competition,analysis,ReasoningGym/intermediate_integration,Apache-2.0,symbolic_antiderivative,60,symbolic
113
+ rg_simple_integration,construct,competition,analysis,ReasoningGym/simple_integration,Apache-2.0,symbolic_antiderivative_polynomial,60,symbolic
114
+ rlve_antiderivative,construct,competition,analysis,RLVE-Gym/integral,MIT,symbolic_antiderivative,60,
115
+ rlve_bezout_identity,construct,competition,number_theory,RLVE-Gym/bezout_identity,MIT,bezout_coefficients,60,agentic_trivial
116
+ rlve_construct_hack_interval,construct,competition,number_theory,RLVE-Gym/construct_hack_interval,MIT,cf468c_hack_interval,60,
117
+ rlve_crt,construct,competition,number_theory,RLVE-Gym/crt,MIT,generalized_crt,60,agentic_trivial
118
+ rlve_distinct_subset_sum_permutation,construct,competition,combinatorics,RLVE-Gym/distinct_array_permutation,MIT,cf892d_gluttony,60,agentic_trivial
119
+ rlve_even_degree_partition,construct,competition,graph_theory,RLVE-Gym/even_degree_graph_partitioning,MIT,gallai_even_partition,60,agentic_trivial
120
+ rlve_graph_isomorphism,construct,competition,graph_theory,RLVE-Gym/graph_isomorphism,MIT,graph_isomorphism,60,agentic_trivial;np_search
121
+ rlve_grid_parity_construction,construct,competition,algebra,RLVE-Gym/grid_parity_construction,MIT,lights_out_gf2,60,agentic_trivial
122
+ rlve_imp_party,construct,competition,graph_theory,RLVE-Gym/imp_party,MIT,poi_party_clique_3n,60,agentic_trivial
123
+ rlve_integer_linear_system,construct,competition,algebra,RLVE-Gym/gaussian_elimination,MIT,integer_linear_system,60,agentic_trivial
124
+ rlve_largest_convex_polygon,construct,competition,discrete_geometry,RLVE-Gym/largest_convex_polygon,MIT,largest_convex_subset,60,agentic_trivial
125
+ rlve_largest_rectangle,construct,competition,discrete_geometry,RLVE-Gym/largest_rectangle_among_points,MIT,largest_rectangle_in_point_set,60,agentic_trivial
126
+ rlve_matrix_permutation_equivalence,construct,competition,combinatorics,RLVE-Gym/matrix_permutation_equivalence,MIT,binary_matrix_permutation_equivalence,60,agentic_trivial;np_search
127
+ rlve_max_achromatic_coloring,construct,competition,graph_theory,RLVE-Gym/maximum_achromatic_number,MIT,achromatic_number,60,agentic_trivial;np_search
128
+ rlve_maximum_clique,construct,competition,graph_theory,RLVE-Gym/maximum_clique,MIT,maximum_clique,60,agentic_trivial;np_search
129
+ rlve_min_chromatic_number,construct,competition,graph_theory,RLVE-Gym/minimum_chromatic_number,MIT,graph_k_coloring_chromatic,60,agentic_trivial;np_search
130
+ rlve_min_harmonious_coloring,construct,competition,graph_theory,RLVE-Gym/minimum_harmonious_chromatic_number,MIT,harmonious_chromatic_number,60,agentic_trivial;np_search
131
+ rlve_pairwise_sums_recovery,construct,competition,combinatorics,RLVE-Gym/squ_squarks,MIT,poi_squarks_pairwise_sums,60,agentic_trivial
132
+ rlve_pan_solar_panels,construct,competition,number_theory,RLVE-Gym/pan_solar_panels,MIT,poi_solar_panels_max_gcd,60,agentic_trivial
133
+ rlve_polynomial_global_minimum,construct,competition,analysis,RLVE-Gym/polynomial_minimum,MIT,polynomial_global_minimum,60,agentic_trivial
134
+ rlve_polynomial_integer_roots,construct,competition,algebra,RLVE-Gym/polynomial_factorization,MIT,integer_root_factorization,60,agentic_trivial;unique_answer
135
+ rlve_polynomial_interpolation,construct,competition,algebra,RLVE-Gym/polynomial_interpolation,MIT,lagrange_interpolation_integer,60,agentic_trivial;unique_answer
136
+ rlve_prefix_product_permutation,construct,competition,number_theory,RLVE-Gym/prefix_product_mod_distinct_permutation,MIT,cf487c_prefix_product_sequence,60,
137
+ rlve_prefix_sum_permutation,construct,competition,number_theory,RLVE-Gym/prefix_sum_mod_distinct_permutation,MIT,prefix_sum_mod_distinct_permutation,60,
138
+ rlve_prime_graph_coloring,construct,competition,graph_theory,RLVE-Gym/prime_graph_minimum_chromatic_number,MIT,prime_distance_graph_coloring,60,agentic_trivial
139
+ rlve_recursive_sequence_sum,construct,competition,number_theory,RLVE-Gym/recursive_sequence_sum_construction,MIT,linear_recurrence_subset_sum,60,agentic_trivial
140
+ rlve_round_robin_scores,construct,competition,combinatorics,RLVE-Gym/round_robin,MIT,tournament_score_realisation_3_1_0,60,agentic_trivial
141
+ rlve_shortest_path_count,construct,competition,combinatorics,RLVE-Gym/shortest_path_count_construction,MIT,cf388b_shortest_path_count,60,
142
+ rlve_smallest_enclosing_circle,construct,competition,geometry,RLVE-Gym/smallest_circle,MIT,minimum_enclosing_circle,60,agentic_trivial
143
+ rlve_subgraph_isomorphism,construct,competition,graph_theory,RLVE-Gym/subgraph_isomorphism,MIT,induced_subgraph_isomorphism,60,agentic_trivial;np_search
144
+ ae_p14_hausdorff_young,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/14-hausdorff-young,CC-BY-4.0,hausdorff_young_sharp_constant,2,float_objective;known_optimum
145
+ ae_p1_ff_nikodym_2d,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym,CC-BY-4.0,finite_field_nikodym_min_size_d2_q_square,3,best_known_uncertain;exact_integer;finite_fields
146
+ ae_p33_thomson,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/33-thomson-problem,CC-BY-4.0,thomson_problem,2,best_known_uncertain;float_objective
147
+ ae_p35_hexagons_in_hexagon,optimize,research,discrete_geometry,AlphaEvolve-Repository-of-Problems/35-packing-in-a-dilate,CC-BY-4.0,unit_hexagons_in_smallest_hexagon,2,float_objective;packing
148
+ ae_p40_erdos_discrepancy_c3,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/40-erdos-discrepancy,CC-BY-4.0,erdos_discrepancy_C3_max_length,1,best_known_uncertain;exact_integer;np_search;sat_friendly
149
+ ae_p56_szemeredi_trotter,optimize,research,combinatorics,AlphaEvolve-Repository-of-Problems/56-szemeredi-trotter,CC-BY-4.0,point_line_incidences_integer_n_eq_m,3,best_known_uncertain;exact_integer;incidence_geometry
150
+ ae_p57_rudin_l4,optimize,research,analysis,AlphaEvolve-Repository-of-Problems/57-rudin-problem,CC-BY-4.0,flat_spherical_harmonic_l4_l2_s2,3,best_known_uncertain;float_objective
151
+ ale_ahc002,optimize,competition,combinatorics,ALE-Bench/ahc002,CC-BY-ND-4.0,ale_ahc002,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
152
+ ale_ahc004,optimize,competition,combinatorics,ALE-Bench/ahc004,CC-BY-ND-4.0,ale_ahc004,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
153
+ ale_ahc009,optimize,competition,probability,ALE-Bench/ahc009,CC-BY-ND-4.0,ale_ahc009,2,ale_bench;best_known_uncertain;float_objective;license_restricted;np_search
154
+ ale_ahc010,optimize,competition,combinatorics,ALE-Bench/ahc010,CC-BY-ND-4.0,ale_ahc010,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
155
+ ale_ahc011,optimize,competition,combinatorics,ALE-Bench/ahc011,CC-BY-ND-4.0,ale_ahc011,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
156
+ ale_ahc014,optimize,competition,discrete_geometry,ALE-Bench/ahc014,CC-BY-ND-4.0,ale_ahc014,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
157
+ ale_ahc017,optimize,competition,graph_theory,ALE-Bench/ahc017,CC-BY-ND-4.0,ale_ahc017,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
158
+ ale_ahc019,optimize,competition,discrete_geometry,ALE-Bench/ahc019,CC-BY-ND-4.0,ale_ahc019,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
159
+ ale_ahc021,optimize,competition,combinatorics,ALE-Bench/ahc021,CC-BY-ND-4.0,ale_ahc021,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
160
+ ale_ahc024,optimize,competition,graph_theory,ALE-Bench/ahc024,CC-BY-ND-4.0,ale_ahc024,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
161
+ ale_ahc026,optimize,competition,combinatorics,ALE-Bench/ahc026,CC-BY-ND-4.0,ale_ahc026,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
162
+ ale_ahc032,optimize,competition,number_theory,ALE-Bench/ahc032,CC-BY-ND-4.0,ale_ahc032,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
163
+ ale_ahc038,optimize,competition,other,ALE-Bench/ahc038,CC-BY-ND-4.0,ale_ahc038,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
164
+ ale_ahc039,optimize,competition,discrete_geometry,ALE-Bench/ahc039,CC-BY-ND-4.0,ale_ahc039,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
165
+ ale_ahc041,optimize,competition,graph_theory,ALE-Bench/ahc041,CC-BY-ND-4.0,ale_ahc041,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
166
+ ale_ahc042,optimize,competition,combinatorics,ALE-Bench/ahc042,CC-BY-ND-4.0,ale_ahc042,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
167
+ ale_ahc044,optimize,competition,combinatorics,ALE-Bench/ahc044,CC-BY-ND-4.0,ale_ahc044,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
168
+ ale_ahc046,optimize,competition,combinatorics,ALE-Bench/ahc046,CC-BY-ND-4.0,ale_ahc046,2,ale_bench;best_known_uncertain;exact_integer;license_restricted;np_search
169
+ edge_apple_incremental,optimize,competition,combinatorics,EdgeBench/apple_incremental_game (AHC058),CC-BY-4.0,edge_ahc058_apple_incremental,2,best_known_uncertain;derived_from_atcoder;edgebench;exact_integer;np_search
170
+ edge_grid_turing_robot,optimize,competition,combinatorics,EdgeBench/grid_turing_robot (AHC056),CC-BY-4.0,edge_ahc056_grid_turing_robot,2,best_known_uncertain;derived_from_atcoder;edgebench;exact_integer;np_search
171
+ fco_mds,optimize,competition,graph_theory,FrontierCO/MDS (CO-Bench/FrontierCO),MIT,frontierco_mds,15,agentic_trivial;best_known_uncertain;exact_integer;frontierco;known_optimum;np_search;related_training:nppc_dominating_set
172
+ fcs_beacon,optimize,competition,combinatorics,FrontierCS/algorithmic/302,MIT,fcs_beacon_string_penalty,3,best_known_uncertain;exact_integer;frontiercs;np_search
173
+ fcs_digit_grid,optimize,competition,combinatorics,FrontierCS/algorithmic/110,MIT,fcs_digit_grid_8x14,1,exact_integer;frontiercs;np_search
174
+ fcs_gcd_chain,optimize,competition,number_theory,FrontierCS/algorithmic/41,MIT,fcs_gcd_increasing_chain_ksum,3,best_known_uncertain;exact_integer;frontiercs
175
+ fcs_inc_subseq_perm,optimize,competition,combinatorics,FrontierCS/algorithmic/33,MIT,fcs_perm_exact_increasing_subsequences,2,exact_integer;frontiercs
176
+ fcs_inverse_paths,optimize,competition,combinatorics,FrontierCS/algorithmic/58,MIT,fcs_inverse_counting_path,3,best_known_uncertain;exact_integer;frontiercs
177
+ fcs_lis_lds_split,optimize,competition,combinatorics,FrontierCS/algorithmic/227,MIT,fcs_lis_lds_four_split,3,best_known_uncertain;exact_integer;frontiercs
178
+ fcs_max3sat,optimize,competition,combinatorics,FrontierCS/algorithmic/176,MIT,fcs_max3sat,3,best_known_uncertain;exact_integer;frontiercs;np_search
179
+ fcs_maxcut,optimize,competition,graph_theory,FrontierCS/algorithmic/192,MIT,fcs_maxcut,3,best_known_uncertain;exact_integer;frontiercs;np_search
180
+ fcs_mult_partial_sums,optimize,competition,number_theory,FrontierCS/algorithmic/83,MIT,completely_multiplicative_min_partial_sum_n1e6,1,exact_integer;frontiercs;related_training:ae_p40_erdos_discrepancy
181
+ fcs_no_rectangle,optimize,competition,combinatorics,FrontierCS/algorithmic/27,MIT,fcs_no_rectangle,3,best_known_uncertain;exact_integer;frontiercs;np_search;zarankiewicz
182
+ fcs_prime_retune,optimize,competition,combinatorics,FrontierCS/algorithmic/314,MIT,fcs_prime_resonance_retuning,3,best_known_uncertain;exact_integer;frontiercs;np_search
183
+ fcs_shortcut3hop,optimize,competition,graph_theory,FrontierCS/algorithmic/239,MIT,fcs_path_shortcut_hop3,3,exact_integer;frontiercs
184
+ fcs_swap_shift,optimize,competition,combinatorics,FrontierCS/algorithmic/247,MIT,fcs_swap_shift_min_ops,3,exact_integer;frontiercs
185
+ fcs_xor_sidon,optimize,competition,combinatorics,FrontierCS/algorithmic/111,MIT,fcs_xor_sidon,3,best_known_uncertain;exact_integer;frontiercs;np_search
186
+ fsh_abc_quality,optimize,research,number_theory,FrontierShift/fs-0375,unknown,abc_triple_max_quality,1,exact_check;float_objective;frontiershift;license_unknown;record_memorizable
187
+ fsh_boolean_nonlinearity_9,optimize,research,coding_theory,FrontierShift/fs-0243,unknown,boolean_function_max_nonlinearity_n9,1,exact_integer;frontiershift;license_unknown;np_search;reference_below_best_known
188
+ fsh_cramer_gap_ratio,optimize,research,number_theory,FrontierShift/fs-0377,unknown,prime_gap_cramer_shanks_granville_ratio_p_lt_2pow64,1,exact_check;float_objective;frontiershift;known_optimum;license_unknown;record_memorizable
189
+ fsh_lwc_1280,optimize,research,coding_theory,FrontierShift/fs-0364 (decodingchallenge.org LW n=1280 seed 0),unknown,decoding_challenge_low_weight_codeword_n1280_seed0,1,cryptography;exact_check;frontiershift;license_unknown;np_search
190
+ fsh_matmul_rank,optimize,research,algebra,"FrontierShift/fs-0034,fs-0035",unknown,matmul_tensor_rank_nnn_over_Q,2,exact_check;exact_rational;frontiershift;license_unknown;record_memorizable;tensor_decomposition
191
+ fsh_maxdet_23,optimize,research,combinatorics,FrontierShift/fs-0164,unknown,hadamard_maxdet_n23,1,best_known_uncertain;exact_integer;frontiershift;license_unknown;np_search
192
+ fsh_rational_diophantine_7,optimize,research,number_theory,FrontierShift/fs-0395,unknown,rational_diophantine_septuple_square_pairs,1,exact_check;exact_rational;frontiershift;license_unknown;reference_below_best_known
193
+ fsh_tetrahedra_packing,optimize,research,discrete_geometry,FrontierShift/fs-0247,unknown,regular_tetrahedra_periodic_packing_density,1,best_known_is_open_target;exact_check;exact_rational;float_objective;frontiershift;license_unknown;reference_below_best_known
194
+ lib_cage,optimize,research,graph_theory,LeapsInBounds/cage,MIT,cage_order_k_g,8,exact_check;exact_integer;leaps_in_bounds;np_search;reference_below_record
195
+ lib_circle_in_square,optimize,research,discrete_geometry,LeapsInBounds/circle-in-square,MIT,equal_circles_in_square_min_side_n,6,best_known_uncertain;exact_check;float_objective;known_optimum;leaps_in_bounds
196
+ lib_circle_in_triangle,optimize,research,discrete_geometry,LeapsInBounds/circle-in-equilateral,MIT,equal_circles_in_equilateral_triangle_min_side_n,3,exact_check;float_objective;known_optimum;leaps_in_bounds
197
+ lib_coil_in_the_box,optimize,research,graph_theory,LeapsInBounds/coil-in-the-box,MIT,coil_in_the_box_dn,5,exact_check;exact_integer;leaps_in_bounds;np_search
198
+ lib_degree_diameter,optimize,research,graph_theory,LeapsInBounds/degree-diameter,MIT,degree_diameter_max_order_d_k,8,exact_check;exact_integer;leaps_in_bounds;np_search
199
+ lib_disk_covering,optimize,research,discrete_geometry,LeapsInBounds/disk-covering,MIT,unit_disk_covering_by_n_disks_min_radius,5,exact_check;exact_rational;float_objective;leaps_in_bounds;reference_below_record
200
+ lib_golomb_ruler,optimize,research,combinatorics,LeapsInBounds/golomb-ruler,MIT,golomb_ruler_min_length_n,7,exact_check;exact_integer;known_optimum;leaps_in_bounds;np_search
201
+ lib_isosceles_set,optimize,research,discrete_geometry,LeapsInBounds/isosceles-set,MIT,isosceles_set_max_size_rn,2,exact_check;exact_integer;known_optimum;leaps_in_bounds
202
+ lib_no_three_in_line,optimize,research,discrete_geometry,LeapsInBounds/no-three-in-line,MIT,no_three_in_line_grid_n,6,exact_check;exact_integer;known_optimum;leaps_in_bounds;np_search
203
+ lib_snake_in_the_box,optimize,research,graph_theory,LeapsInBounds/snake-in-the-box,MIT,snake_in_the_box_dn,4,exact_check;exact_integer;leaps_in_bounds;np_search
204
+ lib_sorting_network_depth,optimize,research,combinatorics,LeapsInBounds/sorting-network-depth,MIT,sorting_network_depth_n,8,exact_check;exact_integer;leaps_in_bounds;np_search
205
+ lib_sorting_network_size,optimize,research,combinatorics,LeapsInBounds/sorting-network-size,MIT,sorting_network_size_n,8,exact_check;exact_integer;leaps_in_bounds;np_search
206
+ lib_square_packing,optimize,research,discrete_geometry,LeapsInBounds/square-packing,MIT,unit_squares_in_square_min_side_n,8,exact_check;exact_rational;float_objective;leaps_in_bounds;reference_is_baseline
207
+ lib_superpermutation,optimize,research,combinatorics,LeapsInBounds/minimal-superpermutation,MIT,minimal_superpermutation_n,4,exact_check;exact_integer;leaps_in_bounds;reference_below_record
208
+ tao_c15a_matmul_rank_exponent,optimize,research,algebra,TaoOptimizationConstants/15a,Apache-2.0,matmul_exponent_exact_rank_formats_le6,1,best_known_uncertain;exact_check;exact_integer;float_objective;tao_constants
209
+ tao_c16a_brezis_gallouet_wainger,optimize,research,analysis,TaoOptimizationConstants/16a,Apache-2.0,brezis_gallouet_wainger_torus_L_radial_shell_le40000,1,agentic_trivial;float_objective;reference_not_record;tao_constants
210
+ tao_c19a_berry_esseen,optimize,research,probability,TaoOptimizationConstants/19a,Apache-2.0,berry_esseen_lower_bound_lattice_n_le10000,1,agentic_trivial;best_known_asymptotic;float_objective;reference_not_record;tao_constants
211
+ tao_c24a_komlos_discrepancy,optimize,research,combinatorics,TaoOptimizationConstants/24a,Apache-2.0,komlos_discrepancy_lower_bound_m_le40_n_le20,1,exact_check;exact_integer;np_search;reference_not_record;tao_constants
212
+ tao_c26a_boolean_bohnenblust_hille,optimize,research,analysis,TaoOptimizationConstants/26a,Apache-2.0,boolean_bohnenblust_hille_ratio_n_le20,1,best_known_asymptotic;exact_integer;float_objective;reference_not_record;tao_constants
213
+ tao_c27b_biplanar_chromatic,optimize,research,graph_theory,TaoOptimizationConstants/27b,Apache-2.0,biplanar_max_chromatic_number_n_le32,1,exact_check;exact_integer;known_record_small;tao_constants
214
+ tao_c37a_degree_sensitivity,optimize,research,combinatorics,TaoOptimizationConstants/37a,Apache-2.0,degree_sensitivity_exponent_n_le16,1,exact_check;float_objective;tao_constants
215
+ tao_c40a_lehmer_mahler,optimize,research,number_theory,TaoOptimizationConstants/40a,Apache-2.0,lehmer_min_mahler_measure_gt1_deg_le64,1,agentic_trivial;exact_check;float_objective;tao_constants
216
+ tao_c49a_sunflower_free,optimize,research,combinatorics,Tao-Davis-Ivanisvili Optimization Constants/49a (sunflower-free capacity),Apache-2.0,sperner_sunflower_free_family_max_size,2,best_known_uncertain;exact_integer;np_search;tao_constants
217
+ tao_c71a_fei_ratio,optimize,research,combinatorics,Tao-Davis-Ivanisvili Optimization Constants/71a (Fourier Entropy-Influence),Apache-2.0,fei_constant_lower_bound_balanced_amplification,1,exact_check;float_objective;np_search;tao_constants
218
+ tao_c73a_hollow_lattice_width,optimize,research,discrete_geometry,Tao-Davis-Ivanisvili Optimization Constants/73a (flatness constant in dimension 3),Apache-2.0,hollow_convex_body_max_lattice_width_d3,1,exact_check;exact_rational;tao_constants
219
+ tao_c88a_admissible_tuple,optimize,research,number_theory,Tao-Davis-Ivanisvili Optimization Constants/88a (bounded prime gaps),Apache-2.0,narrowest_admissible_k_tuple,4,agentic_trivial;best_known_uncertain;exact_integer;known_optimum;np_search;tao_constants
220
+ tao_c90a_covering_min_modulus,optimize,research,number_theory,Tao-Davis-Ivanisvili Optimization Constants/90a (covering systems),Apache-2.0,distinct_covering_system_min_modulus_lcm_le_1e8,1,exact_integer;np_search;record_outside_answer_class;tao_constants
tasks/optimize-ae-p1-ff-nikodym-2d-q121/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p1-ff-nikodym-2d-q121/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 11, "q": 121, "nonresidue": 2, "direction": "minimize", "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q121/instruction.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Nikodym sets in the plane over the finite field F_121 (q = p^2, p = 11).
2
+
3
+ The field F_q with q = 121 elements is F_p(w) with p = 11 and w^2 = 2 (a non-square mod 11): its elements are
4
+ a + b*w with a, b in {0, ..., 10}, written as the pair [a, b]; addition is componentwise mod 11 and
5
+ (a + b w)(c + d w) = (a c + 2 b d) + (a d + b c) w (mod 11). A point of F_q^2 is a pair of field elements
6
+ [[x_a, x_b], [y_a, y_b]].
7
+
8
+ A set N of points of F_q^2 is a Nikodym set if every point P of F_q^2 lies on a line whose other q - 1 = 120
9
+ points all belong to N: there is a direction v = (v_1, v_2) != (0, 0) in F_q^2 such that P + t*v is in N for
10
+ every t in F_q with t != 0. The point P itself may or may not be in N.
11
+
12
+ Find a Nikodym set with as few points as possible. Because N is large (close to q^2 = 14641 points), you submit
13
+ its complement: the list of points of F_q^2 that are NOT in N.
14
+
15
+ Answer format: {"removed": [[[x_a, x_b], [y_a, y_b]], ...]} -- integers in [0, 10], no point listed twice.
16
+ Score (minimize): |N| = 14641 - (number of removed points). An answer for which some point of F_q^2 has no
17
+ punctured line inside N is invalid.
18
+
19
+ Reward: 0 for an invalid answer; otherwise 0.1 plus up to 0.9 for progress from a trivial baseline toward the best known value.
20
+
21
+ Write your final answer as JSON to `/workdir/answer.json`. The instance data is also in `/workdir/instance.json`. Only `/workdir/answer.json` is graded. There is no internet access.
tasks/optimize-ae-p1-ff-nikodym-2d-q121/solution/solve.sh ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ # oracle: a known valid construction (planted witness or published solution)
3
+ cat > /workdir/answer.json <<'RLMATH_EOF'
4
+ {"removed": [[[0, 0], [0, 0]], [[0, 4], [0, 0]], [[0, 6], [0, 0]], [[0, 8], [0, 0]], [[0, 9], [0, 0]], [[0, 10], [0, 0]], [[1, 0], [0, 1]], [[1, 0], [2, 5]], [[1, 0], [2, 6]], [[1, 0], [3, 0]], [[1, 0], [4, 3]], [[1, 0], [4, 8]], [[1, 0], [7, 3]], [[1, 0], [8, 0]], [[1, 0], [9, 5]], [[1, 0], [9, 6]], [[1, 1], [0, 1]], [[1, 1], [0, 10]], [[1, 1], [2, 5]], [[1, 1], [2, 6]], [[1, 1], [3, 0]], [[1, 1], [4, 3]], [[1, 1], [4, 8]], [[1, 1], [7, 8]], [[1, 1], [9, 5]], [[1, 1], [9, 6]], [[1, 2], [0, 1]], [[1, 2], [0, 10]], [[1, 2], [2, 6]], [[1, 2], [3, 0]], [[1, 2], [8, 0]], [[1, 2], [9, 5]], [[1, 2], [9, 6]], [[1, 3], [0, 1]], [[1, 3], [0, 10]], [[1, 3], [2, 5]], [[1, 3], [3, 0]], [[1, 3], [4, 3]], [[1, 3], [7, 8]], [[1, 3], [8, 0]], [[1, 3], [9, 6]], [[1, 4], [0, 10]], [[1, 4], [2, 5]], [[1, 4], [3, 0]], [[1, 4], [4, 3]], [[1, 4], [4, 8]], [[1, 4], [8, 0]], [[1, 4], [9, 5]], [[1, 4], [9, 6]], [[1, 5], [0, 1]], [[1, 5], [0, 10]], [[1, 5], [2, 5]], [[1, 5], [2, 6]], [[1, 5], [3, 0]], [[1, 5], [4, 3]], [[1, 5], [4, 8]], [[1, 5], [7, 3]], [[1, 5], [9, 5]], [[1, 5], [9, 6]], [[1, 6], [0, 1]], [[1, 6], [0, 10]], [[1, 6], [2, 5]], [[1, 6], [2, 6]], [[1, 6], [3, 0]], [[1, 6], [7, 3]], [[1, 6], [7, 8]], [[1, 6], [8, 0]], [[1, 6], [9, 5]], [[1, 6], [9, 6]], [[1, 7], [0, 1]], [[1, 7], [0, 10]], [[1, 7], [2, 5]], [[1, 7], [2, 6]], [[1, 7], [3, 0]], [[1, 7], [4, 3]], [[1, 7], [4, 8]], [[1, 7], [8, 0]], [[1, 7], [9, 5]], [[1, 8], [0, 1]], [[1, 8], [2, 5]], [[1, 8], [2, 6]], [[1, 8], [3, 0]], [[1, 8], [4, 3]], [[1, 8], [7, 8]], [[1, 8], [8, 0]], [[1, 8], [9, 5]], [[1, 8], [9, 6]], [[1, 9], [0, 1]], [[1, 9], [0, 10]], [[1, 9], [2, 5]], [[1, 9], [3, 0]], [[1, 9], [4, 3]], [[1, 9], [4, 8]], [[1, 9], [7, 3]], [[1, 9], [8, 0]], [[1, 9], [9, 6]], [[1, 10], [0, 10]], [[1, 10], [2, 5]], [[1, 10], [3, 0]], [[1, 10], [4, 3]], [[1, 10], [4, 8]], [[1, 10], [7, 3]], [[1, 10], [9, 5]], [[1, 10], [9, 6]], [[2, 0], [2, 2]], [[2, 0], [3, 10]], [[2, 0], [5, 3]], [[2, 0], [5, 8]], [[2, 0], [6, 8]], [[2, 0], [8, 10]], [[2, 0], [9, 2]], [[2, 1], [2, 2]], [[2, 1], [2, 9]], [[2, 1], [3, 1]], [[2, 1], [5, 8]], [[2, 1], [6, 3]], [[2, 1], [8, 1]], [[2, 1], [8, 10]], [[2, 1], [9, 2]], [[2, 1], [9, 9]], [[2, 2], [2, 2]], [[2, 2], [2, 9]], [[2, 2], [3, 1]], [[2, 2], [3, 10]], [[2, 2], [5, 3]], [[2, 2], [5, 8]], [[2, 2], [6, 3]], [[2, 2], [6, 8]], [[2, 2], [8, 1]], [[2, 2], [9, 2]], [[2, 3], [2, 9]], [[2, 3], [3, 1]], [[2, 3], [3, 10]], [[2, 3], [5, 3]], [[2, 3], [5, 8]], [[2, 3], [6, 3]], [[2, 3], [6, 8]], [[2, 3], [8, 1]], [[2, 3], [9, 2]], [[2, 4], [2, 9]], [[2, 4], [3, 1]], [[2, 4], [3, 10]], [[2, 4], [5, 3]], [[2, 4], [5, 8]], [[2, 4], [8, 1]], [[2, 4], [9, 2]], [[2, 5], [2, 2]], [[2, 5], [2, 9]], [[2, 5], [3, 10]], [[2, 5], [5, 3]], [[2, 5], [6, 3]], [[2, 5], [6, 8]], [[2, 5], [9, 2]], [[2, 5], [9, 9]], [[2, 6], [2, 2]], [[2, 6], [2, 9]], [[2, 6], [3, 10]], [[2, 6], [5, 3]], [[2, 6], [6, 3]], [[2, 6], [8, 10]], [[2, 6], [9, 2]], [[2, 6], [9, 9]], [[2, 7], [2, 2]], [[2, 7], [2, 9]], [[2, 7], [3, 1]], [[2, 7], [3, 10]], [[2, 7], [5, 3]], [[2, 7], [5, 8]], [[2, 7], [6, 8]], [[2, 7], [8, 1]], [[2, 7], [8, 10]], [[2, 7], [9, 2]], [[2, 7], [9, 9]], [[2, 8], [2, 2]], [[2, 8], [2, 9]], [[2, 8], [3, 1]], [[2, 8], [3, 10]], [[2, 8], [5, 3]], [[2, 8], [5, 8]], [[2, 8], [8, 1]], [[2, 8], [8, 10]], [[2, 8], [9, 9]], [[2, 9], [2, 9]], [[2, 9], [3, 1]], [[2, 9], [5, 3]], [[2, 9], [5, 8]], [[2, 9], [6, 3]], [[2, 9], [8, 1]], [[2, 9], [8, 10]], [[2, 9], [9, 2]], [[2, 9], [9, 9]], [[2, 10], [2, 2]], [[2, 10], [3, 10]], [[2, 10], [5, 3]], [[2, 10], [5, 8]], [[2, 10], [6, 3]], [[2, 10], [6, 8]], [[2, 10], [8, 10]], [[2, 10], [9, 2]], [[3, 0], [0, 5]], [[3, 0], [1, 3]], [[3, 0], [1, 8]], [[3, 0], [2, 4]], [[3, 0], [7, 0]], [[3, 0], [9, 4]], [[3, 0], [10, 3]], [[3, 0], [10, 8]], [[3, 1], [0, 6]], [[3, 1], [1, 3]], [[3, 1], [7, 0]], [[3, 1], [10, 3]], [[3, 2], [0, 5]], [[3, 2], [0, 6]], [[3, 2], [1, 3]], [[3, 2], [1, 8]], [[3, 2], [2, 4]], [[3, 2], [4, 0]], [[3, 2], [7, 0]], [[3, 2], [9, 4]], [[3, 2], [9, 7]], [[3, 2], [10, 3]], [[3, 2], [10, 8]], [[3, 3], [0, 5]], [[3, 3], [0, 6]], [[3, 3], [1, 3]], [[3, 3], [1, 8]], [[3, 3], [2, 4]], [[3, 3], [2, 7]], [[3, 3], [4, 0]], [[3, 3], [7, 0]], [[3, 3], [9, 4]], [[3, 3], [10, 3]], [[3, 4], [0, 5]], [[3, 4], [1, 3]], [[3, 4], [1, 8]], [[3, 4], [2, 4]], [[3, 4], [2, 7]], [[3, 4], [4, 0]], [[3, 4], [7, 0]], [[3, 5], [0, 5]], [[3, 5], [0, 6]], [[3, 5], [1, 3]], [[3, 5], [1, 8]], [[3, 5], [2, 7]], [[3, 5], [7, 0]], [[3, 5], [9, 4]], [[3, 5], [9, 7]], [[3, 5], [10, 3]], [[3, 6], [0, 5]], [[3, 6], [0, 6]], [[3, 6], [1, 3]], [[3, 6], [1, 8]], [[3, 6], [2, 7]], [[3, 6], [4, 0]], [[3, 6], [7, 0]], [[3, 6], [9, 4]], [[3, 6], [10, 8]], [[3, 7], [0, 5]], [[3, 7], [0, 6]], [[3, 7], [1, 3]], [[3, 7], [2, 7]], [[3, 7], [4, 0]], [[3, 7], [7, 0]], [[3, 7], [9, 4]], [[3, 7], [9, 7]], [[3, 7], [10, 3]], [[3, 7], [10, 8]], [[3, 8], [0, 5]], [[3, 8], [0, 6]], [[3, 8], [1, 3]], [[3, 8], [1, 8]], [[3, 8], [2, 7]], [[3, 8], [4, 0]], [[3, 8], [7, 0]], [[3, 8], [9, 4]], [[3, 8], [10, 3]], [[3, 8], [10, 8]], [[3, 9], [0, 6]], [[3, 9], [1, 3]], [[3, 9], [1, 8]], [[3, 9], [2, 7]], [[3, 9], [4, 0]], [[3, 9], [7, 0]], [[3, 9], [9, 4]], [[3, 9], [9, 7]], [[3, 9], [10, 8]], [[3, 10], [0, 5]], [[3, 10], [1, 3]], [[3, 10], [1, 8]], [[3, 10], [2, 4]], [[3, 10], [2, 7]], [[3, 10], [4, 0]], [[3, 10], [7, 0]], [[3, 10], [9, 4]], [[3, 10], [10, 3]], [[3, 10], [10, 8]], [[4, 0], [0, 2]], [[4, 0], [0, 9]], [[4, 0], [3, 5]], [[4, 0], [3, 6]], [[4, 0], [4, 10]], [[4, 0], [7, 1]], [[4, 0], [7, 10]], [[4, 1], [0, 2]], [[4, 1], [3, 5]], [[4, 1], [4, 10]], [[4, 1], [5, 0]], [[4, 1], [6, 0]], [[4, 1], [7, 1]], [[4, 1], [7, 10]], [[4, 1], [8, 5]], [[4, 1], [8, 6]], [[4, 2], [3, 5]], [[4, 2], [3, 6]], [[4, 2], [4, 1]], [[4, 2], [4, 10]], [[4, 2], [5, 0]], [[4, 2], [6, 0]], [[4, 2], [7, 1]], [[4, 2], [8, 5]], [[4, 2], [8, 6]], [[4, 3], [0, 9]], [[4, 3], [3, 5]], [[4, 3], [4, 10]], [[4, 3], [6, 0]], [[4, 3], [7, 1]], [[4, 3], [8, 6]], [[4, 4], [0, 2]], [[4, 4], [0, 9]], [[4, 4], [3, 5]], [[4, 4], [3, 6]], [[4, 4], [4, 1]], [[4, 4], [4, 10]], [[4, 4], [6, 0]], [[4, 4], [7, 1]], [[4, 4], [7, 10]], [[4, 4], [8, 6]], [[4, 5], [0, 2]], [[4, 5], [0, 9]], [[4, 5], [3, 5]], [[4, 5], [4, 10]], [[4, 5], [5, 0]], [[4, 5], [6, 0]], [[4, 5], [7, 1]], [[4, 5], [8, 5]], [[4, 5], [8, 6]], [[4, 6], [0, 2]], [[4, 6], [0, 9]], [[4, 6], [3, 5]], [[4, 6], [3, 6]], [[4, 6], [4, 1]], [[4, 6], [4, 10]], [[4, 6], [7, 10]], [[4, 6], [8, 5]], [[4, 7], [0, 2]], [[4, 7], [0, 9]], [[4, 7], [3, 6]], [[4, 7], [4, 1]], [[4, 7], [4, 10]], [[4, 7], [5, 0]], [[4, 7], [6, 0]], [[4, 7], [7, 1]], [[4, 7], [7, 10]], [[4, 7], [8, 5]], [[4, 7], [8, 6]], [[4, 8], [0, 9]], [[4, 8], [3, 6]], [[4, 8], [4, 1]], [[4, 8], [5, 0]], [[4, 8], [6, 0]], [[4, 8], [7, 1]], [[4, 8], [7, 10]], [[4, 8], [8, 6]], [[4, 9], [0, 2]], [[4, 9], [0, 9]], [[4, 9], [3, 5]], [[4, 9], [3, 6]], [[4, 9], [4, 1]], [[4, 9], [4, 10]], [[4, 9], [6, 0]], [[4, 9], [7, 1]], [[4, 9], [8, 6]], [[4, 10], [0, 2]], [[4, 10], [0, 9]], [[4, 10], [4, 1]], [[4, 10], [4, 10]], [[4, 10], [5, 0]], [[4, 10], [6, 0]], [[4, 10], [7, 1]], [[4, 10], [7, 10]], [[4, 10], [8, 5]], [[4, 10], [8, 6]], [[5, 0], [0, 4]], [[5, 0], [0, 7]], [[5, 0], [3, 2]], [[5, 0], [3, 9]], [[5, 0], [5, 1]], [[5, 0], [5, 10]], [[5, 0], [6, 1]], [[5, 0], [6, 10]], [[5, 0], [8, 2]], [[5, 1], [0, 4]], [[5, 1], [1, 0]], [[5, 1], [3, 2]], [[5, 1], [3, 9]], [[5, 1], [5, 1]], [[5, 1], [5, 10]], [[5, 1], [6, 10]], [[5, 1], [8, 2]], [[5, 1], [8, 9]], [[5, 2], [0, 4]], [[5, 2], [0, 7]], [[5, 2], [3, 9]], [[5, 2], [5, 1]], [[5, 2], [5, 10]], [[5, 2], [6, 10]], [[5, 2], [8, 2]], [[5, 2], [10, 0]], [[5, 3], [0, 4]], [[5, 3], [0, 7]], [[5, 3], [1, 0]], [[5, 3], [3, 2]], [[5, 3], [3, 9]], [[5, 3], [5, 1]], [[5, 3], [6, 1]], [[5, 3], [6, 10]], [[5, 3], [8, 2]], [[5, 3], [8, 9]], [[5, 3], [10, 0]], [[5, 4], [0, 4]], [[5, 4], [1, 0]], [[5, 4], [3, 2]], [[5, 4], [3, 9]], [[5, 4], [5, 1]], [[5, 4], [5, 10]], [[5, 4], [6, 1]], [[5, 4], [6, 10]], [[5, 4], [8, 2]], [[5, 4], [8, 9]], [[5, 5], [5, 10]], [[5, 5], [6, 1]], [[5, 5], [6, 10]], [[5, 5], [8, 9]], [[5, 5], [10, 0]], [[5, 6], [1, 0]], [[5, 6], [3, 2]], [[5, 6], [3, 9]], [[5, 6], [5, 1]], [[5, 6], [5, 10]], [[5, 6], [6, 1]], [[5, 6], [6, 10]], [[5, 6], [10, 0]], [[5, 7], [0, 4]], [[5, 7], [0, 7]], [[5, 7], [1, 0]], [[5, 7], [3, 2]], [[5, 7], [5, 1]], [[5, 7], [5, 10]], [[5, 7], [6, 1]], [[5, 7], [8, 9]], [[5, 8], [0, 4]], [[5, 8], [1, 0]], [[5, 8], [3, 9]], [[5, 8], [5, 1]], [[5, 8], [5, 10]], [[5, 8], [6, 1]], [[5, 8], [6, 10]], [[5, 8], [8, 2]], [[5, 8], [8, 9]], [[5, 8], [10, 0]], [[5, 9], [0, 4]], [[5, 9], [3, 2]], [[5, 9], [5, 1]], [[5, 9], [6, 10]], [[5, 10], [0, 7]], [[5, 10], [3, 2]], [[5, 10], [3, 9]], [[5, 10], [5, 1]], [[5, 10], [5, 10]], [[5, 10], [6, 1]], [[5, 10], [6, 10]], [[5, 10], [8, 2]], [[5, 10], [10, 0]], [[6, 0], [1, 1]], [[6, 0], [1, 10]], [[6, 0], [3, 4]], [[6, 0], [3, 7]], [[6, 0], [4, 5]], [[6, 0], [7, 5]], [[6, 0], [7, 6]], [[6, 0], [8, 4]], [[6, 0], [8, 7]], [[6, 0], [10, 1]], [[6, 1], [1, 10]], [[6, 1], [3, 4]], [[6, 1], [3, 7]], [[6, 1], [4, 5]], [[6, 1], [4, 6]], [[6, 1], [7, 6]], [[6, 1], [8, 4]], [[6, 1], [10, 1]], [[6, 1], [10, 10]], [[6, 2], [1, 10]], [[6, 2], [3, 4]], [[6, 2], [3, 7]], [[6, 2], [4, 5]], [[6, 2], [7, 5]], [[6, 2], [7, 6]], [[6, 2], [8, 4]], [[6, 2], [10, 1]], [[6, 2], [10, 10]], [[6, 3], [1, 1]], [[6, 3], [1, 10]], [[6, 3], [3, 4]], [[6, 3], [3, 7]], [[6, 3], [4, 5]], [[6, 3], [4, 6]], [[6, 3], [7, 6]], [[6, 3], [8, 7]], [[6, 3], [10, 10]], [[6, 4], [1, 10]], [[6, 4], [3, 7]], [[6, 4], [4, 5]], [[6, 4], [4, 6]], [[6, 4], [8, 4]], [[6, 4], [10, 1]], [[6, 4], [10, 10]], [[6, 5], [1, 1]], [[6, 5], [1, 10]], [[6, 5], [3, 4]], [[6, 5], [3, 7]], [[6, 5], [4, 5]], [[6, 5], [4, 6]], [[6, 5], [8, 4]], [[6, 5], [8, 7]], [[6, 6], [1, 1]], [[6, 6], [1, 10]], [[6, 6], [3, 4]], [[6, 6], [3, 7]], [[6, 6], [4, 5]], [[6, 6], [4, 6]], [[6, 6], [7, 5]], [[6, 6], [8, 4]], [[6, 6], [10, 1]], [[6, 6], [10, 10]], [[6, 7], [1, 1]], [[6, 7], [1, 10]], [[6, 7], [3, 4]], [[6, 7], [3, 7]], [[6, 7], [4, 5]], [[6, 7], [7, 5]], [[6, 7], [7, 6]], [[6, 7], [8, 7]], [[6, 7], [10, 1]], [[6, 7], [10, 10]], [[6, 8], [1, 1]], [[6, 8], [1, 10]], [[6, 8], [3, 4]], [[6, 8], [3, 7]], [[6, 8], [4, 5]], [[6, 8], [4, 6]], [[6, 8], [7, 5]], [[6, 8], [10, 10]], [[6, 9], [1, 1]], [[6, 9], [1, 10]], [[6, 9], [3, 4]], [[6, 9], [3, 7]], [[6, 9], [4, 5]], [[6, 9], [7, 5]], [[6, 9], [8, 4]], [[6, 9], [10, 1]], [[6, 9], [10, 10]], [[6, 10], [1, 1]], [[6, 10], [1, 10]], [[6, 10], [3, 4]], [[6, 10], [4, 5]], [[6, 10], [4, 6]], [[6, 10], [7, 5]], [[6, 10], [7, 6]], [[6, 10], [10, 10]], [[7, 0], [2, 3]], [[7, 0], [2, 8]], [[7, 0], [4, 9]], [[7, 0], [6, 6]], [[7, 0], [9, 8]], [[7, 1], [2, 3]], [[7, 1], [4, 9]], [[7, 1], [5, 5]], [[7, 1], [5, 6]], [[7, 1], [6, 5]], [[7, 1], [7, 2]], [[7, 1], [7, 9]], [[7, 1], [9, 3]], [[7, 2], [2, 3]], [[7, 2], [2, 8]], [[7, 2], [4, 9]], [[7, 2], [5, 5]], [[7, 2], [5, 6]], [[7, 2], [6, 5]], [[7, 2], [7, 2]], [[7, 2], [9, 8]], [[7, 3], [2, 3]], [[7, 3], [2, 8]], [[7, 3], [4, 9]], [[7, 3], [5, 5]], [[7, 3], [6, 5]], [[7, 3], [6, 6]], [[7, 3], [7, 2]], [[7, 3], [7, 9]], [[7, 3], [9, 3]], [[7, 3], [9, 8]], [[7, 4], [2, 3]], [[7, 4], [2, 8]], [[7, 4], [4, 9]], [[7, 4], [5, 5]], [[7, 4], [5, 6]], [[7, 4], [6, 5]], [[7, 4], [6, 6]], [[7, 4], [7, 2]], [[7, 4], [7, 9]], [[7, 4], [9, 8]], [[7, 5], [2, 3]], [[7, 5], [4, 9]], [[7, 5], [5, 5]], [[7, 5], [5, 6]], [[7, 5], [6, 6]], [[7, 5], [7, 9]], [[7, 5], [9, 3]], [[7, 5], [9, 8]], [[7, 6], [2, 3]], [[7, 6], [2, 8]], [[7, 6], [4, 9]], [[7, 6], [5, 5]], [[7, 6], [5, 6]], [[7, 6], [6, 5]], [[7, 6], [6, 6]], [[7, 6], [7, 2]], [[7, 6], [9, 8]], [[7, 7], [2, 3]], [[7, 7], [2, 8]], [[7, 7], [6, 6]], [[7, 7], [7, 2]], [[7, 7], [7, 9]], [[7, 7], [9, 3]], [[7, 7], [9, 8]], [[7, 8], [2, 3]], [[7, 8], [2, 8]], [[7, 8], [4, 9]], [[7, 8], [5, 6]], [[7, 8], [7, 2]], [[7, 8], [9, 3]], [[7, 8], [9, 8]], [[7, 9], [2, 8]], [[7, 9], [4, 9]], [[7, 9], [5, 5]], [[7, 9], [5, 6]], [[7, 9], [6, 5]], [[7, 9], [6, 6]], [[7, 9], [7, 9]], [[7, 9], [9, 3]], [[7, 9], [9, 8]], [[7, 10], [2, 3]], [[7, 10], [4, 9]], [[7, 10], [5, 5]], [[7, 10], [5, 6]], [[7, 10], [6, 5]], [[7, 10], [6, 6]], [[7, 10], [7, 2]], [[7, 10], [9, 3]], [[8, 0], [1, 5]], [[8, 0], [1, 6]], [[8, 0], [4, 4]], [[8, 0], [4, 7]], [[8, 0], [5, 9]], [[8, 0], [6, 2]], [[8, 0], [6, 9]], [[8, 0], [7, 4]], [[8, 0], [7, 7]], [[8, 0], [10, 5]], [[8, 0], [10, 6]], [[8, 1], [1, 5]], [[8, 1], [1, 6]], [[8, 1], [4, 4]], [[8, 1], [4, 7]], [[8, 1], [5, 2]], [[8, 1], [5, 9]], [[8, 1], [6, 2]], [[8, 1], [6, 9]], [[8, 1], [7, 7]], [[8, 1], [10, 5]], [[8, 1], [10, 6]], [[8, 2], [1, 6]], [[8, 2], [4, 4]], [[8, 2], [5, 9]], [[8, 2], [6, 2]], [[8, 2], [6, 9]], [[8, 2], [7, 4]], [[8, 2], [7, 7]], [[8, 2], [10, 5]], [[8, 3], [4, 4]], [[8, 3], [4, 7]], [[8, 3], [5, 2]], [[8, 3], [5, 9]], [[8, 3], [6, 2]], [[8, 3], [7, 4]], [[8, 3], [7, 7]], [[8, 3], [10, 5]], [[8, 4], [1, 5]], [[8, 4], [4, 4]], [[8, 4], [4, 7]], [[8, 4], [5, 2]], [[8, 4], [5, 9]], [[8, 4], [6, 2]], [[8, 4], [6, 9]], [[8, 4], [7, 4]], [[8, 4], [7, 7]], [[8, 4], [10, 5]], [[8, 4], [10, 6]], [[8, 5], [1, 5]], [[8, 5], [1, 6]], [[8, 5], [4, 4]], [[8, 5], [4, 7]], [[8, 5], [5, 2]], [[8, 5], [5, 9]], [[8, 5], [6, 2]], [[8, 5], [7, 4]], [[8, 5], [10, 5]], [[8, 5], [10, 6]], [[8, 6], [1, 6]], [[8, 6], [4, 4]], [[8, 6], [4, 7]], [[8, 6], [5, 2]], [[8, 6], [5, 9]], [[8, 6], [6, 2]], [[8, 6], [6, 9]], [[8, 6], [10, 5]], [[8, 6], [10, 6]], [[8, 7], [1, 5]], [[8, 7], [4, 4]], [[8, 7], [4, 7]], [[8, 7], [5, 2]], [[8, 7], [5, 9]], [[8, 7], [6, 2]], [[8, 7], [7, 4]], [[8, 7], [7, 7]], [[8, 7], [10, 5]], [[8, 7], [10, 6]], [[8, 8], [1, 5]], [[8, 8], [4, 4]], [[8, 8], [5, 2]], [[8, 8], [5, 9]], [[8, 8], [6, 2]], [[8, 8], [6, 9]], [[8, 8], [7, 4]], [[8, 8], [7, 7]], [[8, 8], [10, 5]], [[8, 8], [10, 6]], [[8, 9], [1, 6]], [[8, 9], [4, 4]], [[8, 9], [4, 7]], [[8, 9], [5, 2]], [[8, 9], [5, 9]], [[8, 9], [6, 2]], [[8, 9], [6, 9]], [[8, 9], [7, 4]], [[8, 9], [7, 7]], [[8, 9], [10, 5]], [[8, 9], [10, 6]], [[8, 10], [1, 6]], [[8, 10], [4, 4]], [[8, 10], [4, 7]], [[8, 10], [5, 2]], [[8, 10], [5, 9]], [[8, 10], [6, 2]], [[8, 10], [6, 9]], [[8, 10], [7, 4]], [[8, 10], [7, 7]], [[8, 10], [10, 6]], [[9, 0], [0, 3]], [[9, 0], [0, 8]], [[9, 0], [1, 2]], [[9, 0], [1, 9]], [[9, 0], [2, 0]], [[9, 0], [5, 7]], [[9, 0], [6, 4]], [[9, 0], [6, 7]], [[9, 0], [9, 0]], [[9, 0], [10, 2]], [[9, 0], [10, 9]], [[9, 1], [0, 3]], [[9, 1], [1, 2]], [[9, 1], [1, 9]], [[9, 1], [5, 4]], [[9, 1], [5, 7]], [[9, 1], [6, 4]], [[9, 1], [9, 0]], [[9, 1], [10, 2]], [[9, 1], [10, 9]], [[9, 2], [0, 3]], [[9, 2], [0, 8]], [[9, 2], [1, 2]], [[9, 2], [5, 7]], [[9, 2], [6, 4]], [[9, 2], [6, 7]], [[9, 2], [9, 0]], [[9, 2], [10, 2]], [[9, 2], [10, 9]], [[9, 3], [0, 3]], [[9, 3], [0, 8]], [[9, 3], [1, 2]], [[9, 3], [1, 9]], [[9, 3], [2, 0]], [[9, 3], [5, 4]], [[9, 3], [6, 4]], [[9, 3], [6, 7]], [[9, 3], [9, 0]], [[9, 3], [10, 9]], [[9, 4], [0, 3]], [[9, 4], [0, 8]], [[9, 4], [1, 2]], [[9, 4], [5, 4]], [[9, 4], [5, 7]], [[9, 4], [6, 4]], [[9, 4], [9, 0]], [[9, 5], [0, 3]], [[9, 5], [1, 9]], [[9, 5], [2, 0]], [[9, 5], [5, 4]], [[9, 5], [5, 7]], [[9, 5], [6, 4]], [[9, 5], [9, 0]], [[9, 5], [10, 2]], [[9, 5], [10, 9]], [[9, 6], [0, 3]], [[9, 6], [0, 8]], [[9, 6], [2, 0]], [[9, 6], [5, 4]], [[9, 6], [5, 7]], [[9, 6], [6, 4]], [[9, 6], [6, 7]], [[9, 6], [9, 0]], [[9, 6], [10, 2]], [[9, 6], [10, 9]], [[9, 7], [0, 3]], [[9, 7], [0, 8]], [[9, 7], [2, 0]], [[9, 7], [5, 4]], [[9, 7], [5, 7]], [[9, 7], [6, 4]], [[9, 7], [6, 7]], [[9, 7], [9, 0]], [[9, 7], [10, 2]], [[9, 7], [10, 9]], [[9, 8], [0, 3]], [[9, 8], [0, 8]], [[9, 8], [1, 9]], [[9, 8], [2, 0]], [[9, 8], [5, 4]], [[9, 8], [6, 4]], [[9, 8], [6, 7]], [[9, 8], [9, 0]], [[9, 8], [10, 2]], [[9, 8], [10, 9]], [[9, 9], [0, 3]], [[9, 9], [1, 2]], [[9, 9], [1, 9]], [[9, 9], [2, 0]], [[9, 9], [5, 4]], [[9, 9], [5, 7]], [[9, 9], [6, 4]], [[9, 9], [9, 0]], [[9, 9], [10, 9]], [[9, 10], [0, 3]], [[9, 10], [1, 2]], [[9, 10], [1, 9]], [[9, 10], [2, 0]], [[9, 10], [5, 4]], [[9, 10], [5, 7]], [[9, 10], [10, 9]], [[10, 0], [1, 4]], [[10, 0], [1, 7]], [[10, 0], [2, 1]], [[10, 0], [2, 10]], [[10, 0], [3, 3]], [[10, 0], [8, 3]], [[10, 0], [8, 8]], [[10, 0], [9, 1]], [[10, 0], [9, 10]], [[10, 0], [10, 4]], [[10, 0], [10, 7]], [[10, 1], [1, 4]], [[10, 1], [1, 7]], [[10, 1], [2, 1]], [[10, 1], [2, 10]], [[10, 1], [3, 3]], [[10, 1], [3, 8]], [[10, 1], [8, 8]], [[10, 1], [9, 1]], [[10, 1], [10, 4]], [[10, 2], [1, 4]], [[10, 2], [2, 1]], [[10, 2], [3, 3]], [[10, 2], [8, 8]], [[10, 2], [9, 10]], [[10, 2], [10, 4]], [[10, 3], [2, 1]], [[10, 3], [2, 10]], [[10, 3], [3, 3]], [[10, 3], [3, 8]], [[10, 3], [8, 8]], [[10, 3], [9, 10]], [[10, 3], [10, 4]], [[10, 3], [10, 7]], [[10, 4], [2, 10]], [[10, 4], [3, 3]], [[10, 4], [3, 8]], [[10, 4], [8, 3]], [[10, 4], [8, 8]], [[10, 4], [9, 1]], [[10, 4], [9, 10]], [[10, 4], [10, 4]], [[10, 5], [1, 4]], [[10, 5], [2, 1]], [[10, 5], [3, 8]], [[10, 5], [8, 3]], [[10, 5], [9, 1]], [[10, 5], [9, 10]], [[10, 5], [10, 4]], [[10, 5], [10, 7]], [[10, 6], [1, 4]], [[10, 6], [1, 7]], [[10, 6], [2, 1]], [[10, 6], [2, 10]], [[10, 6], [3, 3]], [[10, 6], [8, 3]], [[10, 6], [8, 8]], [[10, 6], [10, 4]], [[10, 6], [10, 7]], [[10, 7], [1, 4]], [[10, 7], [1, 7]], [[10, 7], [2, 1]], [[10, 7], [2, 10]], [[10, 7], [3, 3]], [[10, 7], [8, 8]], [[10, 7], [9, 1]], [[10, 7], [9, 10]], [[10, 7], [10, 4]], [[10, 7], [10, 7]], [[10, 8], [1, 4]], [[10, 8], [1, 7]], [[10, 8], [2, 1]], [[10, 8], [2, 10]], [[10, 8], [3, 3]], [[10, 8], [3, 8]], [[10, 8], [8, 3]], [[10, 8], [8, 8]], [[10, 8], [9, 10]], [[10, 8], [10, 4]], [[10, 8], [10, 7]], [[10, 9], [1, 4]], [[10, 9], [1, 7]], [[10, 9], [2, 1]], [[10, 9], [2, 10]], [[10, 9], [3, 3]], [[10, 9], [3, 8]], [[10, 9], [8, 3]], [[10, 9], [8, 8]], [[10, 9], [9, 1]], [[10, 9], [9, 10]], [[10, 9], [10, 7]], [[10, 10], [1, 4]], [[10, 10], [1, 7]], [[10, 10], [2, 1]], [[10, 10], [2, 10]], [[10, 10], [3, 3]], [[10, 10], [8, 3]], [[10, 10], [8, 8]], [[10, 10], [9, 10]], [[10, 10], [10, 4]], [[10, 10], [10, 7]]]}
5
+ RLMATH_EOF
tasks/optimize-ae-p1-ff-nikodym-2d-q121/task.toml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version = "1.4"
2
+ artifacts = []
3
+
4
+ [task]
5
+ name = "rlmath/optimize-ae-p1-ff-nikodym-2d-q121"
6
+ version = "0.2.0"
7
+ description = "optimize/ae_p1_ff_nikodym_2d"
8
+ authors = []
9
+ keywords = ["rlmath", "optimize", "ae_p1_ff_nikodym_2d"]
10
+
11
+ [metadata]
12
+ subset = "optimize"
13
+ family = "ae_p1_ff_nikodym_2d"
14
+ problem_key = "finite_field_nikodym_min_size_d2_q_square"
15
+ domain = "combinatorics"
16
+ tier = "research"
17
+ level = "q=121"
18
+ source = "AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym"
19
+ source_license = "CC-BY-4.0"
20
+ tags = ["best_known_uncertain", "exact_integer", "finite_fields"]
21
+ direction = "minimize"
22
+ baseline = 14641
23
+ best_known = 13661
24
+
25
+ [verifier]
26
+ timeout_sec = 300.0
27
+
28
+ [verifier.env]
29
+
30
+ [agent]
31
+ timeout_sec = 900.0
32
+
33
+ [environment]
34
+ network_mode = "no-network"
35
+ build_timeout_sec = 600.0
36
+ os = "linux"
37
+ cpus = 2
38
+ memory_mb = 4096
39
+ mcp_servers = []
40
+
41
+ [environment.env]
42
+
43
+ [solution.env]
tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/check.py ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Finite-field Nikodym set N in F_q^2, q = p^2, submitted as its complement. Score = |N| (minimize).
2
+
3
+ F_q = F_p(w) with w^2 = k (k = inst["nonresidue"], the least quadratic non-residue mod p); the element a + b w
4
+ is written [a, b]. A point is [[x_a, x_b], [y_a, y_b]]. Every point P of F_q^2 needs a direction v != 0 such
5
+ that P + t v lies in N for all t in F_q, t != 0 (P itself may lie outside N). All q + 1 directions
6
+ (1, m), m in F_q, and (0, 1) are checked, vectorised over all q^2 points: a line {P + t v} is labelled by
7
+ the invariant v_y * x - v_x * y in F_q. Upstream (finite_field_nikodym.ipynb, 2D experiments) returned N
8
+ itself as an array; here the complement list must be duplicate-free and in range.
9
+ """
10
+ import numpy as np
11
+
12
+ from run import Invalid, int_list
13
+
14
+
15
+ def check(inst, ans):
16
+ p, k = inst["p"], inst["nonresidue"]
17
+ q = p * p
18
+ rem = ans.get("removed")
19
+ if not isinstance(rem, list):
20
+ raise Invalid("removed must be a list")
21
+ if len(rem) >= q * q:
22
+ raise Invalid("cannot remove every point")
23
+ codes = []
24
+ for pt in rem:
25
+ if not isinstance(pt, list) or len(pt) != 2:
26
+ raise Invalid("each point must be [[x_a, x_b], [y_a, y_b]]")
27
+ (xa, xb), (ya, yb) = int_list(pt[0], 2, 0, p - 1, "x"), int_list(pt[1], 2, 0, p - 1, "y")
28
+ codes.append(((xa * p + xb) * p + ya) * p + yb)
29
+ codes = np.array(codes, dtype=np.int64)
30
+ if len(np.unique(codes)) != len(codes):
31
+ raise Invalid("duplicate removed points")
32
+ in_n = np.ones(q * q, dtype=bool)
33
+ in_n[codes] = False
34
+ idx = np.arange(q * q)
35
+ xa, xb, ya, yb = idx // (p ** 3), (idx // (p * p)) % p, (idx // p) % p, idx % p
36
+ need = q - 1 + in_n.astype(np.int64) # points of N on the line through P, other than P
37
+ good = np.zeros(q * q, dtype=bool)
38
+ # direction (0, 1): invariant = x
39
+ dirs = [None] + [(ma, mb) for ma in range(p) for mb in range(p)]
40
+ for d in dirs:
41
+ if d is None:
42
+ lid = xa * p + xb
43
+ else: # direction (1, m): invariant = m*x - y
44
+ ma, mb = d
45
+ ra = (ma * xa + k * mb * xb - ya) % p
46
+ rb = (ma * xb + mb * xa - yb) % p
47
+ lid = ra * p + rb
48
+ cnt = np.bincount(lid[in_n], minlength=q)
49
+ good |= cnt[lid] == need
50
+ if good.all():
51
+ break
52
+ if not good.all():
53
+ i = int(np.flatnonzero(~good)[0])
54
+ raise Invalid(f"point {[[int(xa[i]), int(xb[i])], [int(ya[i]), int(yb[i])]]} lies on no punctured "
55
+ f"line inside N")
56
+ size = q * q - len(rem)
57
+ return size, {"size": size}
tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 11, "q": 121, "nonresidue": 2, "direction": "minimize", "best_known": 13661, "best_known_source": "AlphaEvolve 2D construction (finite_field_nikodym.ipynb, 2025), re-run offline; size p^4 - p^3 + O(p^2 log p)", "baseline": 14641, "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/run.py ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """rlmath runtime grader. Copied into every task as /tests/run.py next to /tests/check.py.
2
+
3
+ Deterministic; no LLM in the reward path. Runs with python3 + numpy/scipy/sympy/networkx only.
4
+
5
+ python3 /tests/run.py /tests/instance.json /workdir/answer.json
6
+
7
+ Prints the reward in [0, 1] and writes /logs/verifier/details.json when that directory exists.
8
+
9
+ check.py must define check(inst, ans) -> (value, extra_dict) and raise Invalid for a bad answer.
10
+ construct : value is ignored; a returned check means the witness is valid -> reward 1.0
11
+ optimize : value is the raw objective; reward = 0.1 + 0.9 * progress, where
12
+ progress = clip((s - baseline) / (best_known - baseline), 0, 1) (maximize)
13
+ clip((baseline - s) / (baseline - best_known), 0, 1) (minimize)
14
+ inst must carry "direction", "baseline", "best_known".
15
+ """
16
+ import json
17
+ import math
18
+ import os
19
+ import sys
20
+
21
+ VALID_FLOOR = 0.1
22
+
23
+
24
+ class Invalid(Exception):
25
+ pass
26
+
27
+
28
+ # ------------------------------------------------------------ input helpers
29
+
30
+ def int_list(v, n=None, lo=None, hi=None, name="list"):
31
+ if not isinstance(v, list):
32
+ raise Invalid(f"{name} must be a list")
33
+ if n is not None and len(v) != n:
34
+ raise Invalid(f"{name} must have length {n}, got {len(v)}")
35
+ for x in v:
36
+ if isinstance(x, bool) or not isinstance(x, int):
37
+ raise Invalid(f"{name} must contain integers only")
38
+ if (lo is not None and x < lo) or (hi is not None and x > hi):
39
+ raise Invalid(f"{name} entry {x} out of range [{lo}, {hi}]")
40
+ return v
41
+
42
+
43
+ def as_int(x, name="value", lo=None, hi=None):
44
+ return int_list([x], 1, lo, hi, name)[0]
45
+
46
+
47
+ def num(x, name="value"):
48
+ if isinstance(x, bool) or not isinstance(x, (int, float)):
49
+ raise Invalid(f"{name} must be a number")
50
+ if not math.isfinite(x):
51
+ raise Invalid(f"{name} is not finite")
52
+ return float(x)
53
+
54
+
55
+ def float_matrix(v, rows=None, cols=None, name="matrix"):
56
+ if not isinstance(v, list) or (rows is not None and len(v) != rows):
57
+ raise Invalid(f"{name} must be a list of {rows} rows")
58
+ out = []
59
+ for r in v:
60
+ if not isinstance(r, list) or (cols is not None and len(r) != cols):
61
+ raise Invalid(f"each row of {name} must have {cols} numbers")
62
+ out.append([num(x, name) for x in r])
63
+ return out
64
+
65
+
66
+ def float_list(v, n=None, name="list"):
67
+ if not isinstance(v, list) or (n is not None and len(v) != n):
68
+ raise Invalid(f"{name} must be a list of length {n}")
69
+ return [num(x, name) for x in v]
70
+
71
+
72
+ # ------------------------------------------------------------------ reward
73
+
74
+ def optimize_reward(inst, s):
75
+ b, best = inst["baseline"], inst["best_known"]
76
+ if inst.get("progress_scale") == "log":
77
+ # for non-negative losses spanning orders of magnitude (e.g. overlap loss -> 0);
78
+ # progress is measured in log10(value + floor)
79
+ f = inst.get("log_floor", 1e-3)
80
+ b, best, s = (math.log10(max(x, 0.0) + f) for x in (b, best, s))
81
+ if inst["direction"] == "maximize":
82
+ p = (s - b) / (best - b)
83
+ else:
84
+ p = (b - s) / (b - best)
85
+ return VALID_FLOOR + (1 - VALID_FLOOR) * max(0.0, min(1.0, p))
86
+
87
+
88
+ def grade(inst, ans, check):
89
+ try:
90
+ if not isinstance(ans, dict):
91
+ raise Invalid("answer must be a JSON object")
92
+ value, extra = check(inst, ans)
93
+ if inst.get("subset") == "optimize":
94
+ s = float(value)
95
+ if not math.isfinite(s):
96
+ raise Invalid("non-finite score")
97
+ return optimize_reward(inst, s), {"valid": True, "raw_score": s, **(extra or {})}
98
+ return 1.0, {"valid": True, **(extra or {})}
99
+ except Invalid as e:
100
+ return 0.0, {"valid": False, "error": str(e)}
101
+ except Exception as e: # any malformed answer
102
+ return 0.0, {"valid": False, "error": f"{type(e).__name__}: {e}"}
103
+
104
+
105
+ def main():
106
+ here = os.path.dirname(os.path.abspath(__file__))
107
+ sys.path.insert(0, here)
108
+ # import ourselves as `run` so check.py's `from run import Invalid` is the same class
109
+ import run as R # noqa: E402
110
+ import check as family # noqa: E402
111
+
112
+ inst = json.load(open(sys.argv[1]))
113
+ try:
114
+ ans = json.load(open(sys.argv[2]))
115
+ except Exception as e:
116
+ reward, det = 0.0, {"valid": False, "error": f"unreadable answer: {e}"}
117
+ else:
118
+ reward, det = R.grade(inst, ans, family.check)
119
+ det["reward"] = reward
120
+ if os.path.isdir("/logs/verifier"):
121
+ with open("/logs/verifier/details.json", "w") as f:
122
+ json.dump(det, f)
123
+ print(reward)
124
+
125
+
126
+ if __name__ == "__main__":
127
+ main()
tasks/optimize-ae-p1-ff-nikodym-2d-q121/tests/test.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p /logs/verifier
3
+ if [ ! -s /workdir/answer.json ]; then
4
+ echo '{"valid": false, "error": "no /workdir/answer.json"}' > /logs/verifier/details.json
5
+ echo 0 > /logs/verifier/reward.txt
6
+ exit 0
7
+ fi
8
+ timeout 290 python3 /tests/run.py /tests/instance.json /workdir/answer.json \
9
+ > /logs/verifier/reward.txt 2> /logs/verifier/grader.log || echo 0 > /logs/verifier/reward.txt
tasks/optimize-ae-p1-ff-nikodym-2d-q25/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p1-ff-nikodym-2d-q25/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 5, "q": 25, "nonresidue": 2, "direction": "minimize", "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q25/instruction.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Nikodym sets in the plane over the finite field F_25 (q = p^2, p = 5).
2
+
3
+ The field F_q with q = 25 elements is F_p(w) with p = 5 and w^2 = 2 (a non-square mod 5): its elements are
4
+ a + b*w with a, b in {0, ..., 4}, written as the pair [a, b]; addition is componentwise mod 5 and
5
+ (a + b w)(c + d w) = (a c + 2 b d) + (a d + b c) w (mod 5). A point of F_q^2 is a pair of field elements
6
+ [[x_a, x_b], [y_a, y_b]].
7
+
8
+ A set N of points of F_q^2 is a Nikodym set if every point P of F_q^2 lies on a line whose other q - 1 = 24
9
+ points all belong to N: there is a direction v = (v_1, v_2) != (0, 0) in F_q^2 such that P + t*v is in N for
10
+ every t in F_q with t != 0. The point P itself may or may not be in N.
11
+
12
+ Find a Nikodym set with as few points as possible. Because N is large (close to q^2 = 625 points), you submit
13
+ its complement: the list of points of F_q^2 that are NOT in N.
14
+
15
+ Answer format: {"removed": [[[x_a, x_b], [y_a, y_b]], ...]} -- integers in [0, 4], no point listed twice.
16
+ Score (minimize): |N| = 625 - (number of removed points). An answer for which some point of F_q^2 has no
17
+ punctured line inside N is invalid.
18
+
19
+ Reward: 0 for an invalid answer; otherwise 0.1 plus up to 0.9 for progress from a trivial baseline toward the best known value.
20
+
21
+ Write your final answer as JSON to `/workdir/answer.json`. The instance data is also in `/workdir/instance.json`. Only `/workdir/answer.json` is graded. There is no internet access.
tasks/optimize-ae-p1-ff-nikodym-2d-q25/solution/solve.sh ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ # oracle: a known valid construction (planted witness or published solution)
3
+ cat > /workdir/answer.json <<'RLMATH_EOF'
4
+ {"removed": [[[1, 0], [1, 2]], [[1, 0], [1, 3]], [[1, 0], [4, 2]], [[1, 0], [4, 3]], [[1, 1], [1, 2]], [[1, 1], [1, 3]], [[1, 1], [4, 2]], [[1, 1], [4, 3]], [[1, 2], [0, 4]], [[1, 2], [4, 2]], [[1, 2], [4, 3]], [[1, 3], [0, 4]], [[1, 3], [1, 2]], [[1, 3], [4, 2]], [[1, 4], [0, 1]], [[1, 4], [0, 4]], [[1, 4], [1, 3]], [[1, 4], [4, 3]], [[2, 0], [2, 2]], [[2, 0], [2, 3]], [[2, 0], [3, 3]], [[2, 0], [4, 0]], [[2, 1], [2, 2]], [[2, 1], [2, 3]], [[2, 1], [3, 2]], [[2, 1], [3, 3]], [[2, 1], [4, 0]], [[2, 2], [1, 0]], [[2, 2], [3, 3]], [[2, 3], [2, 2]], [[2, 3], [2, 3]], [[2, 3], [3, 2]], [[2, 3], [3, 3]], [[2, 3], [4, 0]], [[2, 4], [1, 0]], [[2, 4], [2, 2]], [[2, 4], [3, 2]], [[2, 4], [3, 3]], [[3, 0], [1, 4]], [[3, 0], [2, 0]], [[3, 0], [3, 0]], [[3, 0], [4, 4]], [[3, 1], [1, 1]], [[3, 1], [2, 0]], [[3, 1], [4, 1]], [[3, 2], [2, 0]], [[3, 2], [3, 0]], [[3, 2], [4, 1]], [[3, 2], [4, 4]], [[3, 3], [1, 1]], [[3, 3], [2, 0]], [[3, 3], [3, 0]], [[3, 3], [4, 1]], [[3, 3], [4, 4]], [[3, 4], [1, 1]], [[3, 4], [1, 4]], [[3, 4], [2, 0]], [[3, 4], [3, 0]], [[3, 4], [4, 1]], [[4, 0], [0, 2]], [[4, 0], [0, 3]], [[4, 0], [2, 1]], [[4, 0], [2, 4]], [[4, 0], [3, 1]], [[4, 1], [0, 2]], [[4, 1], [0, 3]], [[4, 1], [2, 4]], [[4, 2], [0, 2]], [[4, 2], [0, 3]], [[4, 2], [2, 1]], [[4, 2], [3, 4]], [[4, 3], [0, 3]], [[4, 3], [3, 4]], [[4, 4], [0, 2]], [[4, 4], [0, 3]], [[4, 4], [2, 1]], [[4, 4], [3, 1]], [[4, 4], [3, 4]]]}
5
+ RLMATH_EOF
tasks/optimize-ae-p1-ff-nikodym-2d-q25/task.toml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version = "1.4"
2
+ artifacts = []
3
+
4
+ [task]
5
+ name = "rlmath/optimize-ae-p1-ff-nikodym-2d-q25"
6
+ version = "0.2.0"
7
+ description = "optimize/ae_p1_ff_nikodym_2d"
8
+ authors = []
9
+ keywords = ["rlmath", "optimize", "ae_p1_ff_nikodym_2d"]
10
+
11
+ [metadata]
12
+ subset = "optimize"
13
+ family = "ae_p1_ff_nikodym_2d"
14
+ problem_key = "finite_field_nikodym_min_size_d2_q_square"
15
+ domain = "combinatorics"
16
+ tier = "research"
17
+ level = "q=25"
18
+ source = "AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym"
19
+ source_license = "CC-BY-4.0"
20
+ tags = ["best_known_uncertain", "exact_integer", "finite_fields"]
21
+ direction = "minimize"
22
+ baseline = 625
23
+ best_known = 547
24
+
25
+ [verifier]
26
+ timeout_sec = 300.0
27
+
28
+ [verifier.env]
29
+
30
+ [agent]
31
+ timeout_sec = 900.0
32
+
33
+ [environment]
34
+ network_mode = "no-network"
35
+ build_timeout_sec = 600.0
36
+ os = "linux"
37
+ cpus = 2
38
+ memory_mb = 4096
39
+ mcp_servers = []
40
+
41
+ [environment.env]
42
+
43
+ [solution.env]
tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/check.py ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Finite-field Nikodym set N in F_q^2, q = p^2, submitted as its complement. Score = |N| (minimize).
2
+
3
+ F_q = F_p(w) with w^2 = k (k = inst["nonresidue"], the least quadratic non-residue mod p); the element a + b w
4
+ is written [a, b]. A point is [[x_a, x_b], [y_a, y_b]]. Every point P of F_q^2 needs a direction v != 0 such
5
+ that P + t v lies in N for all t in F_q, t != 0 (P itself may lie outside N). All q + 1 directions
6
+ (1, m), m in F_q, and (0, 1) are checked, vectorised over all q^2 points: a line {P + t v} is labelled by
7
+ the invariant v_y * x - v_x * y in F_q. Upstream (finite_field_nikodym.ipynb, 2D experiments) returned N
8
+ itself as an array; here the complement list must be duplicate-free and in range.
9
+ """
10
+ import numpy as np
11
+
12
+ from run import Invalid, int_list
13
+
14
+
15
+ def check(inst, ans):
16
+ p, k = inst["p"], inst["nonresidue"]
17
+ q = p * p
18
+ rem = ans.get("removed")
19
+ if not isinstance(rem, list):
20
+ raise Invalid("removed must be a list")
21
+ if len(rem) >= q * q:
22
+ raise Invalid("cannot remove every point")
23
+ codes = []
24
+ for pt in rem:
25
+ if not isinstance(pt, list) or len(pt) != 2:
26
+ raise Invalid("each point must be [[x_a, x_b], [y_a, y_b]]")
27
+ (xa, xb), (ya, yb) = int_list(pt[0], 2, 0, p - 1, "x"), int_list(pt[1], 2, 0, p - 1, "y")
28
+ codes.append(((xa * p + xb) * p + ya) * p + yb)
29
+ codes = np.array(codes, dtype=np.int64)
30
+ if len(np.unique(codes)) != len(codes):
31
+ raise Invalid("duplicate removed points")
32
+ in_n = np.ones(q * q, dtype=bool)
33
+ in_n[codes] = False
34
+ idx = np.arange(q * q)
35
+ xa, xb, ya, yb = idx // (p ** 3), (idx // (p * p)) % p, (idx // p) % p, idx % p
36
+ need = q - 1 + in_n.astype(np.int64) # points of N on the line through P, other than P
37
+ good = np.zeros(q * q, dtype=bool)
38
+ # direction (0, 1): invariant = x
39
+ dirs = [None] + [(ma, mb) for ma in range(p) for mb in range(p)]
40
+ for d in dirs:
41
+ if d is None:
42
+ lid = xa * p + xb
43
+ else: # direction (1, m): invariant = m*x - y
44
+ ma, mb = d
45
+ ra = (ma * xa + k * mb * xb - ya) % p
46
+ rb = (ma * xb + mb * xa - yb) % p
47
+ lid = ra * p + rb
48
+ cnt = np.bincount(lid[in_n], minlength=q)
49
+ good |= cnt[lid] == need
50
+ if good.all():
51
+ break
52
+ if not good.all():
53
+ i = int(np.flatnonzero(~good)[0])
54
+ raise Invalid(f"point {[[int(xa[i]), int(xb[i])], [int(ya[i]), int(yb[i])]]} lies on no punctured "
55
+ f"line inside N")
56
+ size = q * q - len(rem)
57
+ return size, {"size": size}
tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 5, "q": 25, "nonresidue": 2, "direction": "minimize", "best_known": 547, "best_known_source": "AlphaEvolve 2D construction (finite_field_nikodym.ipynb, 2025), re-run offline; size p^4 - p^3 + O(p^2 log p)", "baseline": 625, "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/run.py ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """rlmath runtime grader. Copied into every task as /tests/run.py next to /tests/check.py.
2
+
3
+ Deterministic; no LLM in the reward path. Runs with python3 + numpy/scipy/sympy/networkx only.
4
+
5
+ python3 /tests/run.py /tests/instance.json /workdir/answer.json
6
+
7
+ Prints the reward in [0, 1] and writes /logs/verifier/details.json when that directory exists.
8
+
9
+ check.py must define check(inst, ans) -> (value, extra_dict) and raise Invalid for a bad answer.
10
+ construct : value is ignored; a returned check means the witness is valid -> reward 1.0
11
+ optimize : value is the raw objective; reward = 0.1 + 0.9 * progress, where
12
+ progress = clip((s - baseline) / (best_known - baseline), 0, 1) (maximize)
13
+ clip((baseline - s) / (baseline - best_known), 0, 1) (minimize)
14
+ inst must carry "direction", "baseline", "best_known".
15
+ """
16
+ import json
17
+ import math
18
+ import os
19
+ import sys
20
+
21
+ VALID_FLOOR = 0.1
22
+
23
+
24
+ class Invalid(Exception):
25
+ pass
26
+
27
+
28
+ # ------------------------------------------------------------ input helpers
29
+
30
+ def int_list(v, n=None, lo=None, hi=None, name="list"):
31
+ if not isinstance(v, list):
32
+ raise Invalid(f"{name} must be a list")
33
+ if n is not None and len(v) != n:
34
+ raise Invalid(f"{name} must have length {n}, got {len(v)}")
35
+ for x in v:
36
+ if isinstance(x, bool) or not isinstance(x, int):
37
+ raise Invalid(f"{name} must contain integers only")
38
+ if (lo is not None and x < lo) or (hi is not None and x > hi):
39
+ raise Invalid(f"{name} entry {x} out of range [{lo}, {hi}]")
40
+ return v
41
+
42
+
43
+ def as_int(x, name="value", lo=None, hi=None):
44
+ return int_list([x], 1, lo, hi, name)[0]
45
+
46
+
47
+ def num(x, name="value"):
48
+ if isinstance(x, bool) or not isinstance(x, (int, float)):
49
+ raise Invalid(f"{name} must be a number")
50
+ if not math.isfinite(x):
51
+ raise Invalid(f"{name} is not finite")
52
+ return float(x)
53
+
54
+
55
+ def float_matrix(v, rows=None, cols=None, name="matrix"):
56
+ if not isinstance(v, list) or (rows is not None and len(v) != rows):
57
+ raise Invalid(f"{name} must be a list of {rows} rows")
58
+ out = []
59
+ for r in v:
60
+ if not isinstance(r, list) or (cols is not None and len(r) != cols):
61
+ raise Invalid(f"each row of {name} must have {cols} numbers")
62
+ out.append([num(x, name) for x in r])
63
+ return out
64
+
65
+
66
+ def float_list(v, n=None, name="list"):
67
+ if not isinstance(v, list) or (n is not None and len(v) != n):
68
+ raise Invalid(f"{name} must be a list of length {n}")
69
+ return [num(x, name) for x in v]
70
+
71
+
72
+ # ------------------------------------------------------------------ reward
73
+
74
+ def optimize_reward(inst, s):
75
+ b, best = inst["baseline"], inst["best_known"]
76
+ if inst.get("progress_scale") == "log":
77
+ # for non-negative losses spanning orders of magnitude (e.g. overlap loss -> 0);
78
+ # progress is measured in log10(value + floor)
79
+ f = inst.get("log_floor", 1e-3)
80
+ b, best, s = (math.log10(max(x, 0.0) + f) for x in (b, best, s))
81
+ if inst["direction"] == "maximize":
82
+ p = (s - b) / (best - b)
83
+ else:
84
+ p = (b - s) / (b - best)
85
+ return VALID_FLOOR + (1 - VALID_FLOOR) * max(0.0, min(1.0, p))
86
+
87
+
88
+ def grade(inst, ans, check):
89
+ try:
90
+ if not isinstance(ans, dict):
91
+ raise Invalid("answer must be a JSON object")
92
+ value, extra = check(inst, ans)
93
+ if inst.get("subset") == "optimize":
94
+ s = float(value)
95
+ if not math.isfinite(s):
96
+ raise Invalid("non-finite score")
97
+ return optimize_reward(inst, s), {"valid": True, "raw_score": s, **(extra or {})}
98
+ return 1.0, {"valid": True, **(extra or {})}
99
+ except Invalid as e:
100
+ return 0.0, {"valid": False, "error": str(e)}
101
+ except Exception as e: # any malformed answer
102
+ return 0.0, {"valid": False, "error": f"{type(e).__name__}: {e}"}
103
+
104
+
105
+ def main():
106
+ here = os.path.dirname(os.path.abspath(__file__))
107
+ sys.path.insert(0, here)
108
+ # import ourselves as `run` so check.py's `from run import Invalid` is the same class
109
+ import run as R # noqa: E402
110
+ import check as family # noqa: E402
111
+
112
+ inst = json.load(open(sys.argv[1]))
113
+ try:
114
+ ans = json.load(open(sys.argv[2]))
115
+ except Exception as e:
116
+ reward, det = 0.0, {"valid": False, "error": f"unreadable answer: {e}"}
117
+ else:
118
+ reward, det = R.grade(inst, ans, family.check)
119
+ det["reward"] = reward
120
+ if os.path.isdir("/logs/verifier"):
121
+ with open("/logs/verifier/details.json", "w") as f:
122
+ json.dump(det, f)
123
+ print(reward)
124
+
125
+
126
+ if __name__ == "__main__":
127
+ main()
tasks/optimize-ae-p1-ff-nikodym-2d-q25/tests/test.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p /logs/verifier
3
+ if [ ! -s /workdir/answer.json ]; then
4
+ echo '{"valid": false, "error": "no /workdir/answer.json"}' > /logs/verifier/details.json
5
+ echo 0 > /logs/verifier/reward.txt
6
+ exit 0
7
+ fi
8
+ timeout 290 python3 /tests/run.py /tests/instance.json /workdir/answer.json \
9
+ > /logs/verifier/reward.txt 2> /logs/verifier/grader.log || echo 0 > /logs/verifier/reward.txt
tasks/optimize-ae-p1-ff-nikodym-2d-q49/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p1-ff-nikodym-2d-q49/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 7, "q": 49, "nonresidue": 3, "direction": "minimize", "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q49/instruction.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Nikodym sets in the plane over the finite field F_49 (q = p^2, p = 7).
2
+
3
+ The field F_q with q = 49 elements is F_p(w) with p = 7 and w^2 = 3 (a non-square mod 7): its elements are
4
+ a + b*w with a, b in {0, ..., 6}, written as the pair [a, b]; addition is componentwise mod 7 and
5
+ (a + b w)(c + d w) = (a c + 3 b d) + (a d + b c) w (mod 7). A point of F_q^2 is a pair of field elements
6
+ [[x_a, x_b], [y_a, y_b]].
7
+
8
+ A set N of points of F_q^2 is a Nikodym set if every point P of F_q^2 lies on a line whose other q - 1 = 48
9
+ points all belong to N: there is a direction v = (v_1, v_2) != (0, 0) in F_q^2 such that P + t*v is in N for
10
+ every t in F_q with t != 0. The point P itself may or may not be in N.
11
+
12
+ Find a Nikodym set with as few points as possible. Because N is large (close to q^2 = 2401 points), you submit
13
+ its complement: the list of points of F_q^2 that are NOT in N.
14
+
15
+ Answer format: {"removed": [[[x_a, x_b], [y_a, y_b]], ...]} -- integers in [0, 6], no point listed twice.
16
+ Score (minimize): |N| = 2401 - (number of removed points). An answer for which some point of F_q^2 has no
17
+ punctured line inside N is invalid.
18
+
19
+ Reward: 0 for an invalid answer; otherwise 0.1 plus up to 0.9 for progress from a trivial baseline toward the best known value.
20
+
21
+ Write your final answer as JSON to `/workdir/answer.json`. The instance data is also in `/workdir/instance.json`. Only `/workdir/answer.json` is graded. There is no internet access.
tasks/optimize-ae-p1-ff-nikodym-2d-q49/solution/solve.sh ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ # oracle: a known valid construction (planted witness or published solution)
3
+ cat > /workdir/answer.json <<'RLMATH_EOF'
4
+ {"removed": [[[0, 0], [0, 0]], [[0, 1], [0, 0]], [[0, 2], [0, 0]], [[0, 3], [0, 0]], [[1, 0], [1, 1]], [[1, 0], [1, 6]], [[1, 0], [2, 3]], [[1, 0], [5, 3]], [[1, 0], [5, 4]], [[1, 0], [6, 1]], [[1, 0], [6, 6]], [[1, 1], [1, 1]], [[1, 1], [2, 3]], [[1, 1], [2, 4]], [[1, 1], [6, 1]], [[1, 1], [6, 6]], [[1, 2], [1, 6]], [[1, 2], [5, 4]], [[1, 2], [6, 1]], [[1, 3], [1, 6]], [[1, 3], [2, 3]], [[1, 3], [5, 3]], [[1, 3], [5, 4]], [[1, 3], [6, 1]], [[1, 3], [6, 6]], [[1, 4], [1, 1]], [[1, 4], [2, 3]], [[1, 4], [5, 3]], [[1, 4], [6, 1]], [[1, 4], [6, 6]], [[1, 5], [1, 1]], [[1, 5], [1, 6]], [[1, 5], [2, 3]], [[1, 5], [2, 4]], [[1, 5], [5, 3]], [[1, 6], [1, 1]], [[1, 6], [1, 6]], [[1, 6], [2, 4]], [[1, 6], [5, 3]], [[1, 6], [5, 4]], [[1, 6], [6, 1]], [[2, 0], [1, 2]], [[2, 0], [1, 5]], [[2, 0], [3, 3]], [[2, 0], [3, 4]], [[2, 0], [4, 3]], [[2, 0], [6, 2]], [[2, 1], [1, 2]], [[2, 1], [3, 3]], [[2, 1], [3, 4]], [[2, 1], [4, 3]], [[2, 1], [4, 4]], [[2, 1], [6, 2]], [[2, 2], [1, 5]], [[2, 2], [3, 3]], [[2, 2], [3, 4]], [[2, 2], [4, 3]], [[2, 2], [6, 2]], [[2, 2], [6, 5]], [[2, 3], [1, 2]], [[2, 3], [1, 5]], [[2, 3], [3, 3]], [[2, 3], [4, 3]], [[2, 3], [4, 4]], [[2, 3], [6, 2]], [[2, 3], [6, 5]], [[2, 4], [1, 2]], [[2, 4], [3, 3]], [[2, 4], [3, 4]], [[2, 4], [4, 3]], [[2, 4], [4, 4]], [[2, 4], [6, 2]], [[2, 4], [6, 5]], [[2, 5], [1, 2]], [[2, 5], [1, 5]], [[2, 5], [3, 4]], [[2, 5], [4, 3]], [[2, 5], [4, 4]], [[2, 5], [6, 2]], [[2, 5], [6, 5]], [[2, 6], [1, 2]], [[2, 6], [1, 5]], [[2, 6], [3, 3]], [[2, 6], [3, 4]], [[2, 6], [4, 3]], [[3, 0], [0, 4]], [[3, 0], [1, 0]], [[3, 0], [2, 6]], [[3, 0], [6, 0]], [[3, 1], [0, 3]], [[3, 1], [1, 0]], [[3, 1], [2, 1]], [[3, 1], [2, 6]], [[3, 1], [5, 1]], [[3, 2], [0, 3]], [[3, 2], [2, 6]], [[3, 2], [5, 1]], [[3, 2], [5, 6]], [[3, 2], [6, 0]], [[3, 3], [0, 3]], [[3, 3], [0, 4]], [[3, 3], [1, 0]], [[3, 3], [2, 1]], [[3, 3], [2, 6]], [[3, 3], [5, 6]], [[3, 4], [0, 3]], [[3, 4], [0, 4]], [[3, 4], [1, 0]], [[3, 4], [2, 1]], [[3, 4], [5, 1]], [[3, 4], [5, 6]], [[3, 5], [0, 3]], [[3, 5], [0, 4]], [[3, 5], [2, 1]], [[3, 5], [5, 6]], [[3, 5], [6, 0]], [[3, 6], [0, 3]], [[3, 6], [0, 4]], [[3, 6], [1, 0]], [[3, 6], [2, 1]], [[3, 6], [2, 6]], [[3, 6], [5, 6]], [[3, 6], [6, 0]], [[4, 0], [2, 2]], [[4, 0], [2, 5]], [[4, 0], [3, 1]], [[4, 0], [3, 6]], [[4, 0], [4, 1]], [[4, 0], [5, 2]], [[4, 0], [5, 5]], [[4, 1], [2, 5]], [[4, 1], [3, 1]], [[4, 1], [3, 6]], [[4, 1], [5, 2]], [[4, 2], [2, 2]], [[4, 2], [3, 1]], [[4, 2], [4, 1]], [[4, 2], [4, 6]], [[4, 3], [2, 2]], [[4, 3], [3, 1]], [[4, 3], [3, 6]], [[4, 3], [5, 2]], [[4, 3], [5, 5]], [[4, 4], [2, 2]], [[4, 4], [2, 5]], [[4, 4], [3, 1]], [[4, 4], [3, 6]], [[4, 5], [2, 2]], [[4, 5], [3, 1]], [[4, 5], [3, 6]], [[4, 5], [5, 2]], [[4, 6], [2, 2]], [[4, 6], [3, 1]], [[4, 6], [3, 6]], [[4, 6], [5, 5]], [[5, 0], [0, 1]], [[5, 0], [0, 6]], [[5, 0], [2, 0]], [[5, 0], [3, 2]], [[5, 0], [3, 5]], [[5, 0], [4, 2]], [[5, 0], [5, 0]], [[5, 1], [0, 6]], [[5, 1], [3, 2]], [[5, 1], [3, 5]], [[5, 1], [4, 2]], [[5, 1], [5, 0]], [[5, 2], [0, 1]], [[5, 2], [0, 6]], [[5, 2], [3, 2]], [[5, 2], [3, 5]], [[5, 2], [4, 2]], [[5, 2], [5, 0]], [[5, 3], [0, 1]], [[5, 3], [3, 2]], [[5, 3], [3, 5]], [[5, 3], [4, 2]], [[5, 3], [4, 5]], [[5, 4], [0, 1]], [[5, 4], [0, 6]], [[5, 4], [2, 0]], [[5, 4], [3, 5]], [[5, 4], [4, 2]], [[5, 4], [4, 5]], [[5, 5], [0, 1]], [[5, 5], [0, 6]], [[5, 5], [3, 2]], [[5, 5], [4, 2]], [[5, 5], [4, 5]], [[5, 5], [5, 0]], [[5, 6], [0, 6]], [[5, 6], [2, 0]], [[5, 6], [3, 2]], [[5, 6], [3, 5]], [[5, 6], [4, 2]], [[5, 6], [4, 5]], [[5, 6], [5, 0]], [[6, 0], [0, 2]], [[6, 0], [0, 5]], [[6, 0], [1, 4]], [[6, 0], [3, 0]], [[6, 0], [4, 0]], [[6, 0], [6, 3]], [[6, 0], [6, 4]], [[6, 1], [0, 5]], [[6, 1], [3, 0]], [[6, 2], [0, 2]], [[6, 2], [0, 5]], [[6, 2], [1, 4]], [[6, 2], [3, 0]], [[6, 2], [6, 3]], [[6, 2], [6, 4]], [[6, 3], [0, 5]], [[6, 3], [1, 4]], [[6, 3], [3, 0]], [[6, 3], [6, 4]], [[6, 4], [0, 2]], [[6, 4], [0, 5]], [[6, 4], [3, 0]], [[6, 4], [4, 0]], [[6, 5], [0, 2]], [[6, 5], [0, 5]], [[6, 5], [3, 0]], [[6, 5], [4, 0]], [[6, 5], [6, 3]], [[6, 5], [6, 4]], [[6, 6], [0, 2]], [[6, 6], [0, 5]], [[6, 6], [1, 4]], [[6, 6], [4, 0]], [[6, 6], [6, 4]]]}
5
+ RLMATH_EOF
tasks/optimize-ae-p1-ff-nikodym-2d-q49/task.toml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version = "1.4"
2
+ artifacts = []
3
+
4
+ [task]
5
+ name = "rlmath/optimize-ae-p1-ff-nikodym-2d-q49"
6
+ version = "0.2.0"
7
+ description = "optimize/ae_p1_ff_nikodym_2d"
8
+ authors = []
9
+ keywords = ["rlmath", "optimize", "ae_p1_ff_nikodym_2d"]
10
+
11
+ [metadata]
12
+ subset = "optimize"
13
+ family = "ae_p1_ff_nikodym_2d"
14
+ problem_key = "finite_field_nikodym_min_size_d2_q_square"
15
+ domain = "combinatorics"
16
+ tier = "research"
17
+ level = "q=49"
18
+ source = "AlphaEvolve-Repository-of-Problems/1-finite-field-nikodym"
19
+ source_license = "CC-BY-4.0"
20
+ tags = ["best_known_uncertain", "exact_integer", "finite_fields"]
21
+ direction = "minimize"
22
+ baseline = 2401
23
+ best_known = 2170
24
+
25
+ [verifier]
26
+ timeout_sec = 300.0
27
+
28
+ [verifier.env]
29
+
30
+ [agent]
31
+ timeout_sec = 900.0
32
+
33
+ [environment]
34
+ network_mode = "no-network"
35
+ build_timeout_sec = 600.0
36
+ os = "linux"
37
+ cpus = 2
38
+ memory_mb = 4096
39
+ mcp_servers = []
40
+
41
+ [environment.env]
42
+
43
+ [solution.env]
tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/check.py ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Finite-field Nikodym set N in F_q^2, q = p^2, submitted as its complement. Score = |N| (minimize).
2
+
3
+ F_q = F_p(w) with w^2 = k (k = inst["nonresidue"], the least quadratic non-residue mod p); the element a + b w
4
+ is written [a, b]. A point is [[x_a, x_b], [y_a, y_b]]. Every point P of F_q^2 needs a direction v != 0 such
5
+ that P + t v lies in N for all t in F_q, t != 0 (P itself may lie outside N). All q + 1 directions
6
+ (1, m), m in F_q, and (0, 1) are checked, vectorised over all q^2 points: a line {P + t v} is labelled by
7
+ the invariant v_y * x - v_x * y in F_q. Upstream (finite_field_nikodym.ipynb, 2D experiments) returned N
8
+ itself as an array; here the complement list must be duplicate-free and in range.
9
+ """
10
+ import numpy as np
11
+
12
+ from run import Invalid, int_list
13
+
14
+
15
+ def check(inst, ans):
16
+ p, k = inst["p"], inst["nonresidue"]
17
+ q = p * p
18
+ rem = ans.get("removed")
19
+ if not isinstance(rem, list):
20
+ raise Invalid("removed must be a list")
21
+ if len(rem) >= q * q:
22
+ raise Invalid("cannot remove every point")
23
+ codes = []
24
+ for pt in rem:
25
+ if not isinstance(pt, list) or len(pt) != 2:
26
+ raise Invalid("each point must be [[x_a, x_b], [y_a, y_b]]")
27
+ (xa, xb), (ya, yb) = int_list(pt[0], 2, 0, p - 1, "x"), int_list(pt[1], 2, 0, p - 1, "y")
28
+ codes.append(((xa * p + xb) * p + ya) * p + yb)
29
+ codes = np.array(codes, dtype=np.int64)
30
+ if len(np.unique(codes)) != len(codes):
31
+ raise Invalid("duplicate removed points")
32
+ in_n = np.ones(q * q, dtype=bool)
33
+ in_n[codes] = False
34
+ idx = np.arange(q * q)
35
+ xa, xb, ya, yb = idx // (p ** 3), (idx // (p * p)) % p, (idx // p) % p, idx % p
36
+ need = q - 1 + in_n.astype(np.int64) # points of N on the line through P, other than P
37
+ good = np.zeros(q * q, dtype=bool)
38
+ # direction (0, 1): invariant = x
39
+ dirs = [None] + [(ma, mb) for ma in range(p) for mb in range(p)]
40
+ for d in dirs:
41
+ if d is None:
42
+ lid = xa * p + xb
43
+ else: # direction (1, m): invariant = m*x - y
44
+ ma, mb = d
45
+ ra = (ma * xa + k * mb * xb - ya) % p
46
+ rb = (ma * xb + mb * xa - yb) % p
47
+ lid = ra * p + rb
48
+ cnt = np.bincount(lid[in_n], minlength=q)
49
+ good |= cnt[lid] == need
50
+ if good.all():
51
+ break
52
+ if not good.all():
53
+ i = int(np.flatnonzero(~good)[0])
54
+ raise Invalid(f"point {[[int(xa[i]), int(xb[i])], [int(ya[i]), int(yb[i])]]} lies on no punctured "
55
+ f"line inside N")
56
+ size = q * q - len(rem)
57
+ return size, {"size": size}
tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"p": 7, "q": 49, "nonresidue": 3, "direction": "minimize", "best_known": 2170, "best_known_source": "AlphaEvolve 2D construction (finite_field_nikodym.ipynb, 2025), re-run offline; size p^4 - p^3 + O(p^2 log p)", "baseline": 2401, "family": "ae_p1_ff_nikodym_2d", "subset": "optimize"}
tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/run.py ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """rlmath runtime grader. Copied into every task as /tests/run.py next to /tests/check.py.
2
+
3
+ Deterministic; no LLM in the reward path. Runs with python3 + numpy/scipy/sympy/networkx only.
4
+
5
+ python3 /tests/run.py /tests/instance.json /workdir/answer.json
6
+
7
+ Prints the reward in [0, 1] and writes /logs/verifier/details.json when that directory exists.
8
+
9
+ check.py must define check(inst, ans) -> (value, extra_dict) and raise Invalid for a bad answer.
10
+ construct : value is ignored; a returned check means the witness is valid -> reward 1.0
11
+ optimize : value is the raw objective; reward = 0.1 + 0.9 * progress, where
12
+ progress = clip((s - baseline) / (best_known - baseline), 0, 1) (maximize)
13
+ clip((baseline - s) / (baseline - best_known), 0, 1) (minimize)
14
+ inst must carry "direction", "baseline", "best_known".
15
+ """
16
+ import json
17
+ import math
18
+ import os
19
+ import sys
20
+
21
+ VALID_FLOOR = 0.1
22
+
23
+
24
+ class Invalid(Exception):
25
+ pass
26
+
27
+
28
+ # ------------------------------------------------------------ input helpers
29
+
30
+ def int_list(v, n=None, lo=None, hi=None, name="list"):
31
+ if not isinstance(v, list):
32
+ raise Invalid(f"{name} must be a list")
33
+ if n is not None and len(v) != n:
34
+ raise Invalid(f"{name} must have length {n}, got {len(v)}")
35
+ for x in v:
36
+ if isinstance(x, bool) or not isinstance(x, int):
37
+ raise Invalid(f"{name} must contain integers only")
38
+ if (lo is not None and x < lo) or (hi is not None and x > hi):
39
+ raise Invalid(f"{name} entry {x} out of range [{lo}, {hi}]")
40
+ return v
41
+
42
+
43
+ def as_int(x, name="value", lo=None, hi=None):
44
+ return int_list([x], 1, lo, hi, name)[0]
45
+
46
+
47
+ def num(x, name="value"):
48
+ if isinstance(x, bool) or not isinstance(x, (int, float)):
49
+ raise Invalid(f"{name} must be a number")
50
+ if not math.isfinite(x):
51
+ raise Invalid(f"{name} is not finite")
52
+ return float(x)
53
+
54
+
55
+ def float_matrix(v, rows=None, cols=None, name="matrix"):
56
+ if not isinstance(v, list) or (rows is not None and len(v) != rows):
57
+ raise Invalid(f"{name} must be a list of {rows} rows")
58
+ out = []
59
+ for r in v:
60
+ if not isinstance(r, list) or (cols is not None and len(r) != cols):
61
+ raise Invalid(f"each row of {name} must have {cols} numbers")
62
+ out.append([num(x, name) for x in r])
63
+ return out
64
+
65
+
66
+ def float_list(v, n=None, name="list"):
67
+ if not isinstance(v, list) or (n is not None and len(v) != n):
68
+ raise Invalid(f"{name} must be a list of length {n}")
69
+ return [num(x, name) for x in v]
70
+
71
+
72
+ # ------------------------------------------------------------------ reward
73
+
74
+ def optimize_reward(inst, s):
75
+ b, best = inst["baseline"], inst["best_known"]
76
+ if inst.get("progress_scale") == "log":
77
+ # for non-negative losses spanning orders of magnitude (e.g. overlap loss -> 0);
78
+ # progress is measured in log10(value + floor)
79
+ f = inst.get("log_floor", 1e-3)
80
+ b, best, s = (math.log10(max(x, 0.0) + f) for x in (b, best, s))
81
+ if inst["direction"] == "maximize":
82
+ p = (s - b) / (best - b)
83
+ else:
84
+ p = (b - s) / (b - best)
85
+ return VALID_FLOOR + (1 - VALID_FLOOR) * max(0.0, min(1.0, p))
86
+
87
+
88
+ def grade(inst, ans, check):
89
+ try:
90
+ if not isinstance(ans, dict):
91
+ raise Invalid("answer must be a JSON object")
92
+ value, extra = check(inst, ans)
93
+ if inst.get("subset") == "optimize":
94
+ s = float(value)
95
+ if not math.isfinite(s):
96
+ raise Invalid("non-finite score")
97
+ return optimize_reward(inst, s), {"valid": True, "raw_score": s, **(extra or {})}
98
+ return 1.0, {"valid": True, **(extra or {})}
99
+ except Invalid as e:
100
+ return 0.0, {"valid": False, "error": str(e)}
101
+ except Exception as e: # any malformed answer
102
+ return 0.0, {"valid": False, "error": f"{type(e).__name__}: {e}"}
103
+
104
+
105
+ def main():
106
+ here = os.path.dirname(os.path.abspath(__file__))
107
+ sys.path.insert(0, here)
108
+ # import ourselves as `run` so check.py's `from run import Invalid` is the same class
109
+ import run as R # noqa: E402
110
+ import check as family # noqa: E402
111
+
112
+ inst = json.load(open(sys.argv[1]))
113
+ try:
114
+ ans = json.load(open(sys.argv[2]))
115
+ except Exception as e:
116
+ reward, det = 0.0, {"valid": False, "error": f"unreadable answer: {e}"}
117
+ else:
118
+ reward, det = R.grade(inst, ans, family.check)
119
+ det["reward"] = reward
120
+ if os.path.isdir("/logs/verifier"):
121
+ with open("/logs/verifier/details.json", "w") as f:
122
+ json.dump(det, f)
123
+ print(reward)
124
+
125
+
126
+ if __name__ == "__main__":
127
+ main()
tasks/optimize-ae-p1-ff-nikodym-2d-q49/tests/test.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p /logs/verifier
3
+ if [ ! -s /workdir/answer.json ]; then
4
+ echo '{"valid": false, "error": "no /workdir/answer.json"}' > /logs/verifier/details.json
5
+ echo 0 > /logs/verifier/reward.txt
6
+ exit 0
7
+ fi
8
+ timeout 290 python3 /tests/run.py /tests/instance.json /workdir/answer.json \
9
+ > /logs/verifier/reward.txt 2> /logs/verifier/grader.log || echo 0 > /logs/verifier/reward.txt
tasks/optimize-ae-p14-hausdorff-young-p4-3/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p14-hausdorff-young-p4-3/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"k": 2, "p": 1.3333333333333333, "p_text": "4/3", "max_len": 10000, "direction": "maximize", "family": "ae_p14_hausdorff_young", "subset": "optimize"}
tasks/optimize-ae-p14-hausdorff-young-p4-3/instruction.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Hausdorff-Young inequality: approach the sharp constant, p = 4/3.
2
+
3
+ Fourier transform convention: f^(xi) = integral f(x) e^(-2 pi i x xi) dx. For p = 4/3 and the dual exponent
4
+ p' = p/(p-1) = 4, the Hausdorff-Young inequality states ||f^||_{p'} <= C(p) ||f||_p for all real
5
+ functions f on the real line, with a sharp constant C(p) <= 1. Consider the quotient
6
+ Q(f) = ||f^||_4 / ||f||_p , ||g||_r = ( integral |g(x)|^r dx )^(1/r),
7
+ and make Q as large as possible.
8
+
9
+ Your f must be a step function with unit steps: f = sum_i h_i * 1_[i, i+1), i = 0, 1, ..., with real heights
10
+ h_i (Q is invariant under dilation and scaling of f, so this costs nothing if you use many steps). Then
11
+ ||f^||_4^4 equals the squared L^2 norm of the 2-fold self-convolution f * ... * f (Plancherel), and Q is
12
+ computed exactly (up to float64 rounding) from a finite sum.
13
+
14
+ Constraints: 1 <= len(h) <= 10000; real entries in [-1e6, 1e6], not all zero.
15
+ Answer format: {"heights": [h_0, h_1, ...]}
16
+ Score (maximize): Q(f).
17
+
18
+ Reward: 0 for an invalid answer; otherwise 0.1 plus up to 0.9 for progress from a trivial baseline toward the best known value.
19
+
20
+ Write your final answer as JSON to `/workdir/answer.json`. The instance data is also in `/workdir/instance.json`. Only `/workdir/answer.json` is graded. There is no internet access.
tasks/optimize-ae-p14-hausdorff-young-p4-3/solution/solve.sh ADDED
The diff for this file is too large to render. See raw diff
 
tasks/optimize-ae-p14-hausdorff-young-p4-3/task.toml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version = "1.4"
2
+ artifacts = []
3
+
4
+ [task]
5
+ name = "rlmath/optimize-ae-p14-hausdorff-young-p4-3"
6
+ version = "0.2.0"
7
+ description = "optimize/ae_p14_hausdorff_young"
8
+ authors = []
9
+ keywords = ["rlmath", "optimize", "ae_p14_hausdorff_young"]
10
+
11
+ [metadata]
12
+ subset = "optimize"
13
+ family = "ae_p14_hausdorff_young"
14
+ problem_key = "hausdorff_young_sharp_constant"
15
+ domain = "analysis"
16
+ tier = "research"
17
+ level = "p=4/3"
18
+ source = "AlphaEvolve-Repository-of-Problems/14-hausdorff-young"
19
+ source_license = "CC-BY-4.0"
20
+ tags = ["float_objective", "known_optimum"]
21
+ direction = "maximize"
22
+ baseline = 0.9036020036098448
23
+ best_known = 0.9366870743752481
24
+
25
+ [verifier]
26
+ timeout_sec = 300.0
27
+
28
+ [verifier.env]
29
+
30
+ [agent]
31
+ timeout_sec = 900.0
32
+
33
+ [environment]
34
+ network_mode = "no-network"
35
+ build_timeout_sec = 600.0
36
+ os = "linux"
37
+ cpus = 2
38
+ memory_mb = 4096
39
+ mcp_servers = []
40
+
41
+ [environment.env]
42
+
43
+ [solution.env]
tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/check.py ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Hausdorff-Young quotient ||f^||_{p'} / ||f||_p for real step functions, p' = 2k an even integer.
2
+
3
+ f = sum_i h_i 1_[i, i+1) (unit steps, real heights), Fourier transform f^(xi) = int f(x) e^{-2 pi i x xi} dx.
4
+ For p' = 2k, (f^)^k is the transform of the k-fold self-convolution f^{*k}, so by Plancherel
5
+ ||f^||_{2k}^{2k} = ||f^{*k}||_2^2,
6
+ and f^{*k} = sum_j c_j B_k(x - j) with c = h^{*k} (discrete convolution) and B_k the order-k cardinal B-spline
7
+ (the k-fold convolution of 1_[0,1)). Hence ||f^{*k}||_2^2 = sum_d M_k(d) sum_j c_j c_{j+d} with
8
+ M_k(d) = int B_k(x) B_k(x + d) dx = (centred order-2k B-spline at d): k=2 -> 2/3, 1/6; k=3 -> 11/20, 13/60, 1/120.
9
+ Everything is a finite sum (float64). Upstream (classical_inequalities.ipynb, p = 3/2) truncated the
10
+ Fourier integral to [-R2, R2] and integrated it by Gaussian quadrature on a fixed 200-point grid; for even p'
11
+ the closed form above replaces that.
12
+ """
13
+ import numpy as np
14
+
15
+ from run import Invalid, float_list
16
+
17
+ M = {2: (2.0 / 3.0, 1.0 / 6.0), 3: (11.0 / 20.0, 13.0 / 60.0, 1.0 / 120.0)}
18
+
19
+
20
+ def check(inst, ans):
21
+ k, p = inst["k"], inst["p"]
22
+ h = ans.get("heights")
23
+ if not isinstance(h, list) or not 1 <= len(h) <= inst["max_len"]:
24
+ raise Invalid(f"heights must be a list of 1..{inst['max_len']} numbers")
25
+ h = np.array(float_list(h, None, "heights"))
26
+ a = np.abs(h)
27
+ if a.max() > 1e6:
28
+ raise Invalid("entries must lie in [-1e6, 1e6]")
29
+ if a.max() == 0:
30
+ raise Invalid("f must not be identically zero")
31
+ h = h / a.max()
32
+ c = h
33
+ for _ in range(k - 1):
34
+ c = np.convolve(c, h)
35
+ w = M[k]
36
+ l2 = w[0] * float(np.dot(c, c)) + 2 * sum(w[d] * float(np.dot(c[:-d], c[d:])) for d in range(1, k)
37
+ if d < len(c))
38
+ if not l2 > 0:
39
+ raise Invalid("degenerate function")
40
+ lp = float(np.sum(np.abs(h) ** p))
41
+ q = l2 ** (1.0 / (2 * k)) / lp ** (1.0 / p)
42
+ return q, {"fhat_norm": l2 ** (1.0 / (2 * k)), "f_norm": lp ** (1.0 / p)}
tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"k": 2, "p": 1.3333333333333333, "p_text": "4/3", "max_len": 10000, "direction": "maximize", "best_known": 0.9366870743752481, "best_known_source": "Babenko-Beckner constant (Babenko 1961; Beckner, Ann. Math. 1975), a proved supremum attained only by Gaussians", "baseline": 0.9036020036098448, "family": "ae_p14_hausdorff_young", "subset": "optimize"}
tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/run.py ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """rlmath runtime grader. Copied into every task as /tests/run.py next to /tests/check.py.
2
+
3
+ Deterministic; no LLM in the reward path. Runs with python3 + numpy/scipy/sympy/networkx only.
4
+
5
+ python3 /tests/run.py /tests/instance.json /workdir/answer.json
6
+
7
+ Prints the reward in [0, 1] and writes /logs/verifier/details.json when that directory exists.
8
+
9
+ check.py must define check(inst, ans) -> (value, extra_dict) and raise Invalid for a bad answer.
10
+ construct : value is ignored; a returned check means the witness is valid -> reward 1.0
11
+ optimize : value is the raw objective; reward = 0.1 + 0.9 * progress, where
12
+ progress = clip((s - baseline) / (best_known - baseline), 0, 1) (maximize)
13
+ clip((baseline - s) / (baseline - best_known), 0, 1) (minimize)
14
+ inst must carry "direction", "baseline", "best_known".
15
+ """
16
+ import json
17
+ import math
18
+ import os
19
+ import sys
20
+
21
+ VALID_FLOOR = 0.1
22
+
23
+
24
+ class Invalid(Exception):
25
+ pass
26
+
27
+
28
+ # ------------------------------------------------------------ input helpers
29
+
30
+ def int_list(v, n=None, lo=None, hi=None, name="list"):
31
+ if not isinstance(v, list):
32
+ raise Invalid(f"{name} must be a list")
33
+ if n is not None and len(v) != n:
34
+ raise Invalid(f"{name} must have length {n}, got {len(v)}")
35
+ for x in v:
36
+ if isinstance(x, bool) or not isinstance(x, int):
37
+ raise Invalid(f"{name} must contain integers only")
38
+ if (lo is not None and x < lo) or (hi is not None and x > hi):
39
+ raise Invalid(f"{name} entry {x} out of range [{lo}, {hi}]")
40
+ return v
41
+
42
+
43
+ def as_int(x, name="value", lo=None, hi=None):
44
+ return int_list([x], 1, lo, hi, name)[0]
45
+
46
+
47
+ def num(x, name="value"):
48
+ if isinstance(x, bool) or not isinstance(x, (int, float)):
49
+ raise Invalid(f"{name} must be a number")
50
+ if not math.isfinite(x):
51
+ raise Invalid(f"{name} is not finite")
52
+ return float(x)
53
+
54
+
55
+ def float_matrix(v, rows=None, cols=None, name="matrix"):
56
+ if not isinstance(v, list) or (rows is not None and len(v) != rows):
57
+ raise Invalid(f"{name} must be a list of {rows} rows")
58
+ out = []
59
+ for r in v:
60
+ if not isinstance(r, list) or (cols is not None and len(r) != cols):
61
+ raise Invalid(f"each row of {name} must have {cols} numbers")
62
+ out.append([num(x, name) for x in r])
63
+ return out
64
+
65
+
66
+ def float_list(v, n=None, name="list"):
67
+ if not isinstance(v, list) or (n is not None and len(v) != n):
68
+ raise Invalid(f"{name} must be a list of length {n}")
69
+ return [num(x, name) for x in v]
70
+
71
+
72
+ # ------------------------------------------------------------------ reward
73
+
74
+ def optimize_reward(inst, s):
75
+ b, best = inst["baseline"], inst["best_known"]
76
+ if inst.get("progress_scale") == "log":
77
+ # for non-negative losses spanning orders of magnitude (e.g. overlap loss -> 0);
78
+ # progress is measured in log10(value + floor)
79
+ f = inst.get("log_floor", 1e-3)
80
+ b, best, s = (math.log10(max(x, 0.0) + f) for x in (b, best, s))
81
+ if inst["direction"] == "maximize":
82
+ p = (s - b) / (best - b)
83
+ else:
84
+ p = (b - s) / (b - best)
85
+ return VALID_FLOOR + (1 - VALID_FLOOR) * max(0.0, min(1.0, p))
86
+
87
+
88
+ def grade(inst, ans, check):
89
+ try:
90
+ if not isinstance(ans, dict):
91
+ raise Invalid("answer must be a JSON object")
92
+ value, extra = check(inst, ans)
93
+ if inst.get("subset") == "optimize":
94
+ s = float(value)
95
+ if not math.isfinite(s):
96
+ raise Invalid("non-finite score")
97
+ return optimize_reward(inst, s), {"valid": True, "raw_score": s, **(extra or {})}
98
+ return 1.0, {"valid": True, **(extra or {})}
99
+ except Invalid as e:
100
+ return 0.0, {"valid": False, "error": str(e)}
101
+ except Exception as e: # any malformed answer
102
+ return 0.0, {"valid": False, "error": f"{type(e).__name__}: {e}"}
103
+
104
+
105
+ def main():
106
+ here = os.path.dirname(os.path.abspath(__file__))
107
+ sys.path.insert(0, here)
108
+ # import ourselves as `run` so check.py's `from run import Invalid` is the same class
109
+ import run as R # noqa: E402
110
+ import check as family # noqa: E402
111
+
112
+ inst = json.load(open(sys.argv[1]))
113
+ try:
114
+ ans = json.load(open(sys.argv[2]))
115
+ except Exception as e:
116
+ reward, det = 0.0, {"valid": False, "error": f"unreadable answer: {e}"}
117
+ else:
118
+ reward, det = R.grade(inst, ans, family.check)
119
+ det["reward"] = reward
120
+ if os.path.isdir("/logs/verifier"):
121
+ with open("/logs/verifier/details.json", "w") as f:
122
+ json.dump(det, f)
123
+ print(reward)
124
+
125
+
126
+ if __name__ == "__main__":
127
+ main()
tasks/optimize-ae-p14-hausdorff-young-p4-3/tests/test.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p /logs/verifier
3
+ if [ ! -s /workdir/answer.json ]; then
4
+ echo '{"valid": false, "error": "no /workdir/answer.json"}' > /logs/verifier/details.json
5
+ echo 0 > /logs/verifier/reward.txt
6
+ exit 0
7
+ fi
8
+ timeout 290 python3 /tests/run.py /tests/instance.json /workdir/answer.json \
9
+ > /logs/verifier/reward.txt 2> /logs/verifier/grader.log || echo 0 > /logs/verifier/reward.txt
tasks/optimize-ae-p14-hausdorff-young-p6-5/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p14-hausdorff-young-p6-5/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"k": 3, "p": 1.2, "p_text": "6/5", "max_len": 10000, "direction": "maximize", "family": "ae_p14_hausdorff_young", "subset": "optimize"}
tasks/optimize-ae-p14-hausdorff-young-p6-5/instruction.md ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Hausdorff-Young inequality: approach the sharp constant, p = 6/5.
2
+
3
+ Fourier transform convention: f^(xi) = integral f(x) e^(-2 pi i x xi) dx. For p = 6/5 and the dual exponent
4
+ p' = p/(p-1) = 6, the Hausdorff-Young inequality states ||f^||_{p'} <= C(p) ||f||_p for all real
5
+ functions f on the real line, with a sharp constant C(p) <= 1. Consider the quotient
6
+ Q(f) = ||f^||_6 / ||f||_p , ||g||_r = ( integral |g(x)|^r dx )^(1/r),
7
+ and make Q as large as possible.
8
+
9
+ Your f must be a step function with unit steps: f = sum_i h_i * 1_[i, i+1), i = 0, 1, ..., with real heights
10
+ h_i (Q is invariant under dilation and scaling of f, so this costs nothing if you use many steps). Then
11
+ ||f^||_6^6 equals the squared L^2 norm of the 3-fold self-convolution f * ... * f (Plancherel), and Q is
12
+ computed exactly (up to float64 rounding) from a finite sum.
13
+
14
+ Constraints: 1 <= len(h) <= 10000; real entries in [-1e6, 1e6], not all zero.
15
+ Answer format: {"heights": [h_0, h_1, ...]}
16
+ Score (maximize): Q(f).
17
+
18
+ Reward: 0 for an invalid answer; otherwise 0.1 plus up to 0.9 for progress from a trivial baseline toward the best known value.
19
+
20
+ Write your final answer as JSON to `/workdir/answer.json`. The instance data is also in `/workdir/instance.json`. Only `/workdir/answer.json` is graded. There is no internet access.
tasks/optimize-ae-p14-hausdorff-young-p6-5/solution/solve.sh ADDED
The diff for this file is too large to render. See raw diff
 
tasks/optimize-ae-p14-hausdorff-young-p6-5/task.toml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version = "1.4"
2
+ artifacts = []
3
+
4
+ [task]
5
+ name = "rlmath/optimize-ae-p14-hausdorff-young-p6-5"
6
+ version = "0.2.0"
7
+ description = "optimize/ae_p14_hausdorff_young"
8
+ authors = []
9
+ keywords = ["rlmath", "optimize", "ae_p14_hausdorff_young"]
10
+
11
+ [metadata]
12
+ subset = "optimize"
13
+ family = "ae_p14_hausdorff_young"
14
+ problem_key = "hausdorff_young_sharp_constant"
15
+ domain = "analysis"
16
+ tier = "research"
17
+ level = "p=6/5"
18
+ source = "AlphaEvolve-Repository-of-Problems/14-hausdorff-young"
19
+ source_license = "CC-BY-4.0"
20
+ tags = ["float_objective", "known_optimum"]
21
+ direction = "maximize"
22
+ baseline = 0.9051636706146827
23
+ best_known = 0.9292792681498123
24
+
25
+ [verifier]
26
+ timeout_sec = 300.0
27
+
28
+ [verifier.env]
29
+
30
+ [agent]
31
+ timeout_sec = 900.0
32
+
33
+ [environment]
34
+ network_mode = "no-network"
35
+ build_timeout_sec = 600.0
36
+ os = "linux"
37
+ cpus = 2
38
+ memory_mb = 4096
39
+ mcp_servers = []
40
+
41
+ [environment.env]
42
+
43
+ [solution.env]
tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/check.py ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """Hausdorff-Young quotient ||f^||_{p'} / ||f||_p for real step functions, p' = 2k an even integer.
2
+
3
+ f = sum_i h_i 1_[i, i+1) (unit steps, real heights), Fourier transform f^(xi) = int f(x) e^{-2 pi i x xi} dx.
4
+ For p' = 2k, (f^)^k is the transform of the k-fold self-convolution f^{*k}, so by Plancherel
5
+ ||f^||_{2k}^{2k} = ||f^{*k}||_2^2,
6
+ and f^{*k} = sum_j c_j B_k(x - j) with c = h^{*k} (discrete convolution) and B_k the order-k cardinal B-spline
7
+ (the k-fold convolution of 1_[0,1)). Hence ||f^{*k}||_2^2 = sum_d M_k(d) sum_j c_j c_{j+d} with
8
+ M_k(d) = int B_k(x) B_k(x + d) dx = (centred order-2k B-spline at d): k=2 -> 2/3, 1/6; k=3 -> 11/20, 13/60, 1/120.
9
+ Everything is a finite sum (float64). Upstream (classical_inequalities.ipynb, p = 3/2) truncated the
10
+ Fourier integral to [-R2, R2] and integrated it by Gaussian quadrature on a fixed 200-point grid; for even p'
11
+ the closed form above replaces that.
12
+ """
13
+ import numpy as np
14
+
15
+ from run import Invalid, float_list
16
+
17
+ M = {2: (2.0 / 3.0, 1.0 / 6.0), 3: (11.0 / 20.0, 13.0 / 60.0, 1.0 / 120.0)}
18
+
19
+
20
+ def check(inst, ans):
21
+ k, p = inst["k"], inst["p"]
22
+ h = ans.get("heights")
23
+ if not isinstance(h, list) or not 1 <= len(h) <= inst["max_len"]:
24
+ raise Invalid(f"heights must be a list of 1..{inst['max_len']} numbers")
25
+ h = np.array(float_list(h, None, "heights"))
26
+ a = np.abs(h)
27
+ if a.max() > 1e6:
28
+ raise Invalid("entries must lie in [-1e6, 1e6]")
29
+ if a.max() == 0:
30
+ raise Invalid("f must not be identically zero")
31
+ h = h / a.max()
32
+ c = h
33
+ for _ in range(k - 1):
34
+ c = np.convolve(c, h)
35
+ w = M[k]
36
+ l2 = w[0] * float(np.dot(c, c)) + 2 * sum(w[d] * float(np.dot(c[:-d], c[d:])) for d in range(1, k)
37
+ if d < len(c))
38
+ if not l2 > 0:
39
+ raise Invalid("degenerate function")
40
+ lp = float(np.sum(np.abs(h) ** p))
41
+ q = l2 ** (1.0 / (2 * k)) / lp ** (1.0 / p)
42
+ return q, {"fhat_norm": l2 ** (1.0 / (2 * k)), "f_norm": lp ** (1.0 / p)}
tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"k": 3, "p": 1.2, "p_text": "6/5", "max_len": 10000, "direction": "maximize", "best_known": 0.9292792681498123, "best_known_source": "Babenko-Beckner constant (Babenko 1961; Beckner, Ann. Math. 1975), a proved supremum attained only by Gaussians", "baseline": 0.9051636706146827, "family": "ae_p14_hausdorff_young", "subset": "optimize"}
tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/run.py ADDED
@@ -0,0 +1,127 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """rlmath runtime grader. Copied into every task as /tests/run.py next to /tests/check.py.
2
+
3
+ Deterministic; no LLM in the reward path. Runs with python3 + numpy/scipy/sympy/networkx only.
4
+
5
+ python3 /tests/run.py /tests/instance.json /workdir/answer.json
6
+
7
+ Prints the reward in [0, 1] and writes /logs/verifier/details.json when that directory exists.
8
+
9
+ check.py must define check(inst, ans) -> (value, extra_dict) and raise Invalid for a bad answer.
10
+ construct : value is ignored; a returned check means the witness is valid -> reward 1.0
11
+ optimize : value is the raw objective; reward = 0.1 + 0.9 * progress, where
12
+ progress = clip((s - baseline) / (best_known - baseline), 0, 1) (maximize)
13
+ clip((baseline - s) / (baseline - best_known), 0, 1) (minimize)
14
+ inst must carry "direction", "baseline", "best_known".
15
+ """
16
+ import json
17
+ import math
18
+ import os
19
+ import sys
20
+
21
+ VALID_FLOOR = 0.1
22
+
23
+
24
+ class Invalid(Exception):
25
+ pass
26
+
27
+
28
+ # ------------------------------------------------------------ input helpers
29
+
30
+ def int_list(v, n=None, lo=None, hi=None, name="list"):
31
+ if not isinstance(v, list):
32
+ raise Invalid(f"{name} must be a list")
33
+ if n is not None and len(v) != n:
34
+ raise Invalid(f"{name} must have length {n}, got {len(v)}")
35
+ for x in v:
36
+ if isinstance(x, bool) or not isinstance(x, int):
37
+ raise Invalid(f"{name} must contain integers only")
38
+ if (lo is not None and x < lo) or (hi is not None and x > hi):
39
+ raise Invalid(f"{name} entry {x} out of range [{lo}, {hi}]")
40
+ return v
41
+
42
+
43
+ def as_int(x, name="value", lo=None, hi=None):
44
+ return int_list([x], 1, lo, hi, name)[0]
45
+
46
+
47
+ def num(x, name="value"):
48
+ if isinstance(x, bool) or not isinstance(x, (int, float)):
49
+ raise Invalid(f"{name} must be a number")
50
+ if not math.isfinite(x):
51
+ raise Invalid(f"{name} is not finite")
52
+ return float(x)
53
+
54
+
55
+ def float_matrix(v, rows=None, cols=None, name="matrix"):
56
+ if not isinstance(v, list) or (rows is not None and len(v) != rows):
57
+ raise Invalid(f"{name} must be a list of {rows} rows")
58
+ out = []
59
+ for r in v:
60
+ if not isinstance(r, list) or (cols is not None and len(r) != cols):
61
+ raise Invalid(f"each row of {name} must have {cols} numbers")
62
+ out.append([num(x, name) for x in r])
63
+ return out
64
+
65
+
66
+ def float_list(v, n=None, name="list"):
67
+ if not isinstance(v, list) or (n is not None and len(v) != n):
68
+ raise Invalid(f"{name} must be a list of length {n}")
69
+ return [num(x, name) for x in v]
70
+
71
+
72
+ # ------------------------------------------------------------------ reward
73
+
74
+ def optimize_reward(inst, s):
75
+ b, best = inst["baseline"], inst["best_known"]
76
+ if inst.get("progress_scale") == "log":
77
+ # for non-negative losses spanning orders of magnitude (e.g. overlap loss -> 0);
78
+ # progress is measured in log10(value + floor)
79
+ f = inst.get("log_floor", 1e-3)
80
+ b, best, s = (math.log10(max(x, 0.0) + f) for x in (b, best, s))
81
+ if inst["direction"] == "maximize":
82
+ p = (s - b) / (best - b)
83
+ else:
84
+ p = (b - s) / (b - best)
85
+ return VALID_FLOOR + (1 - VALID_FLOOR) * max(0.0, min(1.0, p))
86
+
87
+
88
+ def grade(inst, ans, check):
89
+ try:
90
+ if not isinstance(ans, dict):
91
+ raise Invalid("answer must be a JSON object")
92
+ value, extra = check(inst, ans)
93
+ if inst.get("subset") == "optimize":
94
+ s = float(value)
95
+ if not math.isfinite(s):
96
+ raise Invalid("non-finite score")
97
+ return optimize_reward(inst, s), {"valid": True, "raw_score": s, **(extra or {})}
98
+ return 1.0, {"valid": True, **(extra or {})}
99
+ except Invalid as e:
100
+ return 0.0, {"valid": False, "error": str(e)}
101
+ except Exception as e: # any malformed answer
102
+ return 0.0, {"valid": False, "error": f"{type(e).__name__}: {e}"}
103
+
104
+
105
+ def main():
106
+ here = os.path.dirname(os.path.abspath(__file__))
107
+ sys.path.insert(0, here)
108
+ # import ourselves as `run` so check.py's `from run import Invalid` is the same class
109
+ import run as R # noqa: E402
110
+ import check as family # noqa: E402
111
+
112
+ inst = json.load(open(sys.argv[1]))
113
+ try:
114
+ ans = json.load(open(sys.argv[2]))
115
+ except Exception as e:
116
+ reward, det = 0.0, {"valid": False, "error": f"unreadable answer: {e}"}
117
+ else:
118
+ reward, det = R.grade(inst, ans, family.check)
119
+ det["reward"] = reward
120
+ if os.path.isdir("/logs/verifier"):
121
+ with open("/logs/verifier/details.json", "w") as f:
122
+ json.dump(det, f)
123
+ print(reward)
124
+
125
+
126
+ if __name__ == "__main__":
127
+ main()
tasks/optimize-ae-p14-hausdorff-young-p6-5/tests/test.sh ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ #!/bin/bash
2
+ mkdir -p /logs/verifier
3
+ if [ ! -s /workdir/answer.json ]; then
4
+ echo '{"valid": false, "error": "no /workdir/answer.json"}' > /logs/verifier/details.json
5
+ echo 0 > /logs/verifier/reward.txt
6
+ exit 0
7
+ fi
8
+ timeout 290 python3 /tests/run.py /tests/instance.json /workdir/answer.json \
9
+ > /logs/verifier/reward.txt 2> /logs/verifier/grader.log || echo 0 > /logs/verifier/reward.txt
tasks/optimize-ae-p33-thomson-n292/environment/Dockerfile ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ FROM python:3.11.16-slim
2
+ RUN apt-get update && apt-get install -y --no-install-recommends tmux curl git procps build-essential \
3
+ && rm -rf /var/lib/apt/lists/*
4
+ RUN pip install --no-cache-dir numpy==2.4.6 scipy==1.17.1 sympy==1.14.0 networkx==3.6.1 \
5
+ python-sat==1.9.dev15 z3-solver==5.1.0.0 PuLP==3.3.2 ortools==9.15.6755
6
+ WORKDIR /workdir
7
+ COPY instance.json /workdir/instance.json
tasks/optimize-ae-p33-thomson-n292/environment/instance.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"n": 292, "direction": "minimize", "family": "ae_p33_thomson", "subset": "optimize"}