| { |
| "checked_at_utc": "2026-07-19T11:19:22.402493Z", |
| "checks": { |
| "current_source_identity_and_hashes": { |
| "checked_author_source_sha256": { |
| "diffres/integators.py": "4cbcdba683cf27532b21acfa9c3295354c52cc2f4494b4063a0459629eb72954", |
| "diffres/resampling.py": "12b33fe23e1f62e18909cbcff9fa45db36ed65f6f943668728c9fe61795aa06e", |
| "diffres/tools.py": "54c60e1ac4f484532f51be57a63aacbdfb7a53ef8cef621495176ea7a843a5df", |
| "experiments/gms/baselines.py": "7b258df0ab4a7ca8eb5c91f6d5efbabbe359e0130db1cb9e4d3263a9241f1d5c", |
| "experiments/gms/diffusion.py": "886496bab31d254ac8030286ea350e48db817bd680cccc8cdda50c58a94ba774", |
| "experiments/rnd_keys.npy": "6e60685445936e0380220f34fc3fd15b5b7720070dc8c8f7f1148575de4ff5df", |
| "experiments/run_gms.sh": "c01a4c88a3ef05d6a5ffb585b9d4233b1b1a9e8c1c46d40360eab44963b3eee5", |
| "experiments/summary/print_gms_errs.py": "1df81cadaaf35ca5b26e8bd1ce6fc19aa4a60a2f7e161075087e518b3a0d17cd" |
| }, |
| "current_commit": "767effe3e755067eb8a04422597fbf37eb8ab754", |
| "current_origin_matches_expected": true, |
| "current_tracked_worktree_clean": true |
| }, |
| "raw_record_integrity_and_aggregate_recomputation": { |
| "diffusion_lower_is_better_wins": { |
| "posterior_mean_residual_squared_l2": 51, |
| "sliced_wasserstein_l1": 54 |
| }, |
| "paired_diffusion_minus_multinomial_95_percent_t_intervals": { |
| "posterior_mean_residual_squared_l2": { |
| "critical_value": 1.9842169515086827, |
| "lower": -0.007123449582480549, |
| "mean": -0.00044225546225615345, |
| "sample_count": 100, |
| "sample_standard_deviation": 0.03367169157155121, |
| "standard_error": 0.003367169157155121, |
| "upper": 0.006238938657968242 |
| }, |
| "sliced_wasserstein_l1": { |
| "critical_value": 1.9842169515086827, |
| "lower": -0.004659334515850839, |
| "mean": -0.0016648057831093993, |
| "sample_count": 100, |
| "sample_standard_deviation": 0.015091740499770325, |
| "standard_error": 0.0015091740499770326, |
| "upper": 0.0013297229496320405 |
| } |
| }, |
| "recomputed_aggregate": { |
| "diffusion_resampling": { |
| "posterior_mean_residual_squared_l2": { |
| "maximum": 0.13674925583503703, |
| "mean": 0.037390775125730545, |
| "minimum": 0.003946685433082616, |
| "population_standard_deviation": 0.029935755902808506 |
| }, |
| "sliced_wasserstein_l1": { |
| "maximum": 0.13845705348834103, |
| "mean": 0.08076861937013251, |
| "minimum": 0.049275986886894586, |
| "population_standard_deviation": 0.021289921970778995 |
| } |
| }, |
| "multinomial_baseline": { |
| "posterior_mean_residual_squared_l2": { |
| "maximum": 0.34224144904965786, |
| "mean": 0.0378330305879867, |
| "minimum": 0.00328472168500068, |
| "population_standard_deviation": 0.044287731038823226 |
| }, |
| "sliced_wasserstein_l1": { |
| "maximum": 0.19381190521839362, |
| "mean": 0.08243342515324191, |
| "minimum": 0.04778862503625366, |
| "population_standard_deviation": 0.02503231412513419 |
| } |
| } |
| }, |
| "recomputed_paired_aggregate": { |
| "posterior_mean_residual_squared_l2": { |
| "maximum": 0.10671040850944102, |
| "mean": -0.00044225546225615345, |
| "minimum": -0.24669590547035306, |
| "population_standard_deviation": 0.03350291009982232 |
| }, |
| "sliced_wasserstein_l1": { |
| "maximum": 0.041779128105039046, |
| "mean": -0.0016648057831093993, |
| "minimum": -0.07544369540487143, |
| "population_standard_deviation": 0.015016092201344656 |
| } |
| }, |
| "record_count": 100 |
| }, |
| "selected_protocol": "passed", |
| "strict_json": "passed" |
| }, |
| "claim_boundary": "This check independently recomputes the recorded aggregates and paired intervals from the 100 raw JSON records while checking the current pinned source identity. It does not rerun the numerical experiment, independently reproduce the paper, or establish an outperformance or significance claim.", |
| "input": { |
| "path": "study/gms-selected-ids-0-99.json", |
| "sha256": "c74851b88b5df723cbfb28edb48be09eb61e7a2dab086088f90d59591b68723b" |
| }, |
| "schema_version": 1, |
| "validation_status": "passed", |
| "validation_type": "post-hoc integrity and aggregate recomputation; no numerical rerun", |
| "validator": { |
| "path": "study/validate_gms_selected_result.py", |
| "platform": { |
| "machine": "arm64", |
| "release": "25.5.0", |
| "system": "Darwin" |
| }, |
| "python_version": "3.13.14 (main, Jun 23 2026, 15:20:43) [Clang 22.1.3 ]", |
| "scipy_version": "1.16.2", |
| "sha256": "80841d7e4395f60bcbf06880a92288ad7bfdd7e022ea60876911277f3b3e69e5" |
| } |
| } |
|
|
