{ "component": "gdn", "gpu": "NVIDIA B300 SXM6 AC", "torch": "2.12.1+cu130", "method": "Identical inputs for both implementations, original judge unchanged, fp64 rebuilt truth.", "rows": [ { "batch_size": 16, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0025729717609002, "max_individual_head_error_ratio": 1.0047603580901656, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.004840669455448137 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000377502995784, "max_individual_head_error_ratio": 1.0022604107694364, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.004832510374899398 } }, { "batch_size": 16, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0026752459212538, "max_individual_head_error_ratio": 1.006318504577656, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.005649392516492693 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000349169295915, "max_individual_head_error_ratio": 1.0022884314945337, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.005645332649465186 } }, { "batch_size": 16, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0025384655853626, "max_individual_head_error_ratio": 1.0066552397632773, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.005356422144122431 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000355531207006, "max_individual_head_error_ratio": 1.0019881834282838, "individual_states_checked": 512, "individual_states_passing_same_bound": 512, "max_state_relative_mae": 0.005350061724333348 } }, { "batch_size": 32, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002415611201314, "max_individual_head_error_ratio": 1.006853307437364, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1023, "max_state_relative_mae": 0.004920333931390794 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000413978369906, "max_individual_head_error_ratio": 1.0027988669732626, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1024, "max_state_relative_mae": 0.004907739415799461 } }, { "batch_size": 32, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0026713480992455, "max_individual_head_error_ratio": 1.0092145516148332, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1024, "max_state_relative_mae": 0.005523096649382196 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001916456690558, "max_individual_head_error_ratio": 1.003512031597695, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1024, "max_state_relative_mae": 0.005512654366444548 } }, { "batch_size": 32, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0026787060162845, "max_individual_head_error_ratio": 1.0059780827327927, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1024, "max_state_relative_mae": 0.0051969578416112445 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001794866601792, "max_individual_head_error_ratio": 1.0017688926870474, "individual_states_checked": 1024, "individual_states_passing_same_bound": 1024, "max_state_relative_mae": 0.005191103082662031 } }, { "batch_size": 64, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0024645508053158, "max_individual_head_error_ratio": 1.006293791648136, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2048, "max_state_relative_mae": 0.004980311590053733 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001507747739824, "max_individual_head_error_ratio": 1.00317946147758, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2048, "max_state_relative_mae": 0.004982245993474239 } }, { "batch_size": 64, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0030193805950687, "max_individual_head_error_ratio": 1.0110133892697506, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2047, "max_state_relative_mae": 0.005625704411960469 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001581497154506, "max_individual_head_error_ratio": 1.0029824282495465, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2048, "max_state_relative_mae": 0.005618331749020674 } }, { "batch_size": 64, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002311357822909, "max_individual_head_error_ratio": 1.005728529102603, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2048, "max_state_relative_mae": 0.0052580455868875995 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001955458139993, "max_individual_head_error_ratio": 1.0020411638185058, "individual_states_checked": 2048, "individual_states_passing_same_bound": 2048, "max_state_relative_mae": 0.005243640297246211 } }, { "batch_size": 128, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0025462621654009, "max_individual_head_error_ratio": 1.0066856899929304, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4095, "max_state_relative_mae": 0.010526152204327946 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0001947303443228, "max_individual_head_error_ratio": 1.004548568460813, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4096, "max_state_relative_mae": 0.0049031079934334565 } }, { "batch_size": 128, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0032445477224305, "max_individual_head_error_ratio": 1.0318646468062047, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4093, "max_state_relative_mae": 0.005562230379007623 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0003710109379027, "max_individual_head_error_ratio": 1.0083343429514107, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4096, "max_state_relative_mae": 0.005556101315462859 } }, { "batch_size": 128, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002261602486124, "max_individual_head_error_ratio": 1.007637345557252, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4096, "max_state_relative_mae": 0.005311109318430315 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0000838112924513, "max_individual_head_error_ratio": 1.002071687917802, "individual_states_checked": 4096, "individual_states_passing_same_bound": 4096, "max_state_relative_mae": 0.005298019176255004 } }, { "batch_size": 256, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0024647582781914, "max_individual_head_error_ratio": 1.0088686111563892, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8192, "max_state_relative_mae": 0.00512573710625055 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000112040546511, "max_individual_head_error_ratio": 1.0044320133454956, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8192, "max_state_relative_mae": 0.00511842051747702 } }, { "batch_size": 256, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0029729218922754, "max_individual_head_error_ratio": 1.0149323482870427, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8186, "max_state_relative_mae": 0.005777002774164632 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.000375479114499, "max_individual_head_error_ratio": 1.007336808073591, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8192, "max_state_relative_mae": 0.005767604784431954 } }, { "batch_size": 256, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002200382049273, "max_individual_head_error_ratio": 1.0063932299199936, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8192, "max_state_relative_mae": 0.005327109310098326 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0000921392755253, "max_individual_head_error_ratio": 1.0022797211172036, "individual_states_checked": 8192, "individual_states_passing_same_bound": 8192, "max_state_relative_mae": 0.005318594943876018 } }, { "batch_size": 512, "family": "structured", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002529822356002, "max_individual_head_error_ratio": 4.056641018280484, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16383, "max_state_relative_mae": 0.00499518405502506 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0000648081757677, "max_individual_head_error_ratio": 1.003933236442185, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16384, "max_state_relative_mae": 0.004983685305136407 } }, { "batch_size": 512, "family": "second_window", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.0029405041432273, "max_individual_head_error_ratio": 1.0194967933832526, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16367, "max_state_relative_mae": 0.005688874402274295 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0002760946222988, "max_individual_head_error_ratio": 1.0044266372078552, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16384, "max_state_relative_mae": 0.005674156145267935 } }, { "batch_size": 512, "family": "random", "baseline": { "original_judge_passed": true, "worst_head_error_ratio": 1.002165149387159, "max_individual_head_error_ratio": 1.007021719926822, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16384, "max_state_relative_mae": 0.005395732953884217 }, "current": { "original_judge_passed": true, "worst_head_error_ratio": 1.0000556209394427, "max_individual_head_error_ratio": 1.0024445311674375, "individual_states_checked": 16384, "individual_states_passing_same_bound": 16384, "max_state_relative_mae": 0.005390848426091049 } } ], "status": "original_benchmark_pass", "global_seed_plus_batch": 20261026, "relative_tolerance": 0.01, "absolute_cap_factor": 0.001, "reference_error_floor_factor": 0.0001, "individual_state_check_is_additional": true, "summary": { "baseline": { "original_cases_passed": 18, "worst_original_benchmark_error_ratio": 1.0032445477224305, "individual_states_checked": 96768, "additional_individual_state_failures": 30, "max_individual_head_error_ratio": 4.056641018280484 }, "current": { "original_cases_passed": 18, "worst_original_benchmark_error_ratio": 1.000413978369906, "individual_states_checked": 96768, "additional_individual_state_failures": 0, "max_individual_head_error_ratio": 1.0083343429514107 } } }