| { |
| "data_coverage": { |
| "total_dirs": 26257, |
| "has_metrics_json": 26220, |
| "has_ken_test": 1502, |
| "has_config": 3199, |
| "has_claim": 7311, |
| "has_any_ap": 10479, |
| "skipped_no_ap": 15778, |
| "skipped_parse_error": 0 |
| }, |
| "analysis_1_post_bugfix_anova": { |
| "n_post_bugfix": 1177, |
| "n_with_arch": 1177, |
| "post_bugfix_anova": { |
| "f_statistic": 81.43813556464495, |
| "p_value": 1.1102230246251565e-16, |
| "df_between": 15, |
| "df_within": 1156, |
| "eta_squared": 0.5137897049062238, |
| "n_groups": 16, |
| "n_total": 1172, |
| "groups_used": { |
| "VJepa2+Zipformer": 554, |
| "DINOv2+Zipformer": 15, |
| "DINOv3-B+Retention": 39, |
| "VJepa2+Hybrid R-M": 62, |
| "VJepa2+BiMamba": 137, |
| "DINOv3-B+BiMamba": 60, |
| "VJepa2+Retention": 71, |
| "Multi-Backbone+Zipformer": 24, |
| "DINOv3-B+Zipformer": 141, |
| "DINOv3-B+Hybrid R-M": 7, |
| "SigLIP2+Retention": 14, |
| "SigLIP2+Hybrid R-M": 8, |
| "SigLIP2+BiMamba": 5, |
| "SigLIP2+Zipformer": 20, |
| "Multi-Backbone+Hybrid R-M": 9, |
| "Multi-Backbone+BiMamba": 6 |
| } |
| }, |
| "full_dataset_anova": { |
| "f_statistic": 86.44657739999107, |
| "p_value": 1.1102230246251565e-16, |
| "df_between": 18, |
| "df_within": 3077, |
| "eta_squared": 0.3358570037933899, |
| "n_groups": 19, |
| "n_total": 3096, |
| "groups_used": { |
| "VJepa2+Zipformer": 1607, |
| "DINOv3-B+Zipformer": 413, |
| "DINOv3-B+Retention": 104, |
| "VJepa2+Retention": 170, |
| "DINOv3-L+Zipformer": 167, |
| "DINOv2+Zipformer": 130, |
| "VJepa2+Hybrid R-M": 68, |
| "VJepa2+BiMamba": 202, |
| "DINOv3-B+BiMamba": 72, |
| "Multi-Backbone+Zipformer": 24, |
| "DINOv2+Retention": 9, |
| "DINOv3-B+Hybrid R-M": 7, |
| "DINOv3-L+Retention": 35, |
| "SigLIP2+Retention": 17, |
| "SigLIP2+Hybrid R-M": 8, |
| "SigLIP2+Zipformer": 43, |
| "SigLIP2+BiMamba": 5, |
| "Multi-Backbone+Hybrid R-M": 9, |
| "Multi-Backbone+BiMamba": 6 |
| } |
| }, |
| "post_bugfix_group_stats": { |
| "VJepa2+BiMamba": { |
| "count": 137, |
| "mean": 0.7265990907064925, |
| "std": 0.24332331850434874, |
| "best": 0.9755711092285307 |
| }, |
| "VJepa2+Retention": { |
| "count": 71, |
| "mean": 0.7262586624995755, |
| "std": 0.24279622758833574, |
| "best": 0.9209791098744063 |
| }, |
| "VJepa2+Zipformer": { |
| "count": 554, |
| "mean": 0.6985367132335054, |
| "std": 0.2813309960333393, |
| "best": 0.9852744878745394 |
| }, |
| "VJepa2+Hybrid R-M": { |
| "count": 62, |
| "mean": 0.6017963106489552, |
| "std": 0.35966898271583086, |
| "best": 0.9616804371250811 |
| }, |
| "Multi-Backbone+Zipformer": { |
| "count": 24, |
| "mean": 0.5312604199513161, |
| "std": 0.08930154194539587, |
| "best": 0.7751574573174227 |
| }, |
| "Multi-Backbone+BiMamba": { |
| "count": 6, |
| "mean": 0.4890058760497445, |
| "std": 0.08883919177954687, |
| "best": 0.6067720585361953 |
| }, |
| "Multi-Backbone+Hybrid R-M": { |
| "count": 9, |
| "mean": 0.48144465586716834, |
| "std": 0.08714630804669826, |
| "best": 0.6018989337183028 |
| }, |
| "SigLIP2+Hybrid R-M": { |
| "count": 8, |
| "mean": 0.4790050494875844, |
| "std": 0.055110344747337676, |
| "best": 0.5744278006812511 |
| }, |
| "SigLIP2+Retention": { |
| "count": 14, |
| "mean": 0.47800743828741626, |
| "std": 0.03028206123303963, |
| "best": 0.522749041297347 |
| }, |
| "DINOv2+Zipformer": { |
| "count": 15, |
| "mean": 0.4510678247731804, |
| "std": 0.07926686884363983, |
| "best": 0.5128758905472057 |
| }, |
| "SigLIP2+BiMamba": { |
| "count": 5, |
| "mean": 0.39956751709145333, |
| "std": 0.20065483809833934, |
| "best": 0.5235996940951071 |
| }, |
| "SigLIP2+Zipformer": { |
| "count": 20, |
| "mean": 0.32455638741174947, |
| "std": 0.24810544499422968, |
| "best": 0.5916297750320977 |
| }, |
| "DINOv3-B+BiMamba": { |
| "count": 60, |
| "mean": 0.11161920017690977, |
| "std": 0.08855937815649899, |
| "best": 0.46703567661423856 |
| }, |
| "DINOv3-B+Zipformer": { |
| "count": 141, |
| "mean": 0.09442922673644218, |
| "std": 0.06877989044418327, |
| "best": 0.5632991082513329 |
| }, |
| "DINOv3-B+Retention": { |
| "count": 39, |
| "mean": 0.08657997411552709, |
| "std": 0.026772170282191105, |
| "best": 0.153112749019836 |
| }, |
| "DINOv3-B+Hybrid R-M": { |
| "count": 7, |
| "mean": 0.08477993213518317, |
| "std": 0.016075515337445768, |
| "best": 0.11104386043996581 |
| } |
| } |
| }, |
| "analysis_2_test_ap_top_configs": { |
| "n_with_test_ap": 1329, |
| "n_with_both": 1329, |
| "top10_by_val": [ |
| { |
| "idea_id": "idea-2ae88c", |
| "val_ap": 1.0, |
| "test_ap": null, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer", |
| "loss_type": "focal_loss" |
| }, |
| { |
| "idea_id": "idea-2360", |
| "val_ap": 0.9989761736049196, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3240", |
| "val_ap": 0.998959471411477, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3212", |
| "val_ap": 0.9989180977907799, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-1543", |
| "val_ap": 0.9989099819243711, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3304", |
| "val_ap": 0.9988873430091092, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3136", |
| "val_ap": 0.9988806194395451, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-2385", |
| "val_ap": 0.9988679394766856, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3459", |
| "val_ap": 0.9987802167437801, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| }, |
| { |
| "idea_id": "idea-3382", |
| "val_ap": 0.9987305004779815, |
| "test_ap": null, |
| "backbone": null, |
| "encoder": null, |
| "loss_type": null |
| } |
| ], |
| "top10_by_test": [ |
| { |
| "idea_id": "idea-2ec818", |
| "test_ap": 0.9245057778280362, |
| "val_ap": 0.9543569112936242, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-220eb6-ht-8", |
| "test_ap": 0.9205753344470232, |
| "val_ap": 0.967569861166418, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-0fbe1e", |
| "test_ap": 0.9202693153074217, |
| "val_ap": 0.9327693595124967, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-02d0b0", |
| "test_ap": 0.9163091968125885, |
| "val_ap": 0.923359054424835, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-1e60f0-ht-6", |
| "test_ap": 0.9162179332224651, |
| "val_ap": 0.9574796380017213, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-338f86", |
| "test_ap": 0.9132446147493699, |
| "val_ap": 0.9485904820162623, |
| "backbone": "VJepa2", |
| "encoder": "Retention" |
| }, |
| { |
| "idea_id": "idea-14f2f9", |
| "test_ap": 0.9055013231270591, |
| "val_ap": 0.909131668538389, |
| "backbone": "VJepa2", |
| "encoder": "Zipformer" |
| }, |
| { |
| "idea_id": "idea-0ab093", |
| "test_ap": 0.9054196697190158, |
| "val_ap": 0.9437194876623506, |
| "backbone": "VJepa2", |
| "encoder": "Hybrid R-M" |
| }, |
| { |
| "idea_id": "idea-2714f2", |
| "test_ap": 0.9016478432286867, |
| "val_ap": 0.9362953990579823, |
| "backbone": "VJepa2", |
| "encoder": "BiMamba" |
| }, |
| { |
| "idea_id": "idea-434131-ht-1", |
| "test_ap": 0.9009977439794165, |
| "val_ap": 0.9051255754047571, |
| "backbone": "VJepa2", |
| "encoder": "Retention" |
| } |
| ], |
| "val_test_correlation": { |
| "spearman_rho": 0.7005664607546771, |
| "p_value": 1.0443337017392531e-196, |
| "n": 1329 |
| }, |
| "vjepa2_test_stats": { |
| "n": 801, |
| "mean": 0.7612305749966171, |
| "std": 0.163757339706533, |
| "ci95": [ |
| 0.7498827703098123, |
| 0.7725783796834218 |
| ], |
| "best": 0.9245057778280362 |
| }, |
| "other_backbones_test_stats": { |
| "n": 528, |
| "mean": 0.24186660697735782, |
| "std": 0.23923213006760027, |
| "ci95": [ |
| 0.22144122105749744, |
| 0.2622919928972182 |
| ], |
| "best": 0.8235769575519882 |
| } |
| }, |
| "analysis_3_convergence": { |
| "full_dataset": { |
| "n_experiments": 10479, |
| "ap_at_n": { |
| "AP@100": 0.9999506254114549, |
| "AP@500": 1.0, |
| "AP@1000": 1.0, |
| "AP@5000": 1.0, |
| "AP@10000": 1.0, |
| "AP@20000": 1.0 |
| }, |
| "power_law_fit": { |
| "power_law_exponent": 1.6788201767221818, |
| "b_coefficient": 3.15268382758459, |
| "best_final": 1.0, |
| "n_points_for_fit": 110 |
| }, |
| "llm_vs_random": { |
| "n_check": 1000, |
| "llm_ap": 1.0, |
| "random_mean": 1.0, |
| "random_std": 0.0, |
| "llm_advantage": 0.0, |
| "llm_percentile": 0.0 |
| }, |
| "best_ap": 1.0, |
| "first_time": "2026-02-16T03:52:30", |
| "last_time": "2026-03-12T03:49:06" |
| }, |
| "post_bugfix": { |
| "n_experiments": 1177, |
| "ap_at_n": { |
| "AP@100": 0.8740212533822255, |
| "AP@500": 0.9756482941319878, |
| "AP@1000": 0.9852744878745394, |
| "AP@5000": 0.9852744878745394, |
| "AP@10000": 0.9852744878745394, |
| "AP@20000": 0.9852744878745394 |
| }, |
| "power_law_fit": { |
| "power_law_exponent": 0.961002566630579, |
| "b_coefficient": 6.3807002791515846, |
| "best_final": 0.9852744878745394, |
| "n_points_for_fit": 587 |
| }, |
| "llm_vs_random": { |
| "n_check": 1000, |
| "llm_ap": 0.9852744878745394, |
| "random_mean": 0.9840230826880076, |
| "random_std": 0.003237322015973951, |
| "llm_advantage": 0.0012514051865317732, |
| "llm_percentile": 13.0 |
| }, |
| "best_ap": 0.9852744878745394, |
| "first_time": "2026-03-06T00:02:43", |
| "last_time": "2026-03-12T03:49:06" |
| }, |
| "test_ap_only": { |
| "n_experiments": 1329, |
| "ap_at_n": { |
| "AP@100": 0.8140566199161836, |
| "AP@500": 0.8851484305948076, |
| "AP@1000": 0.9245057778280362, |
| "AP@5000": 0.9245057778280362, |
| "AP@10000": 0.9245057778280362, |
| "AP@20000": 0.9245057778280362 |
| }, |
| "power_law_fit": { |
| "power_law_exponent": 0.5397795301394496, |
| "b_coefficient": 1.1540588896373332, |
| "best_final": 0.9245057778280362, |
| "n_points_for_fit": 758 |
| }, |
| "llm_vs_random": { |
| "n_check": 1000, |
| "llm_ap": 0.9245057778280362, |
| "random_mean": 0.9234694981854145, |
| "random_std": 0.001931561229878777, |
| "llm_advantage": 0.0010362796426217624, |
| "llm_percentile": 24.0 |
| }, |
| "best_ap": 0.9245057778280362, |
| "first_time": "2026-03-02T16:34:25", |
| "last_time": "2026-03-09T05:12:50" |
| } |
| }, |
| "analysis_4_full_anova": { |
| "n_with_config_and_ap": 3122, |
| "backbone_anova": { |
| "f_statistic": 271.8638603982837, |
| "p_value": 1.1102230246251565e-16, |
| "df_between": 5, |
| "df_within": 3101, |
| "eta_squared": 0.304758294184122, |
| "n_groups": 6, |
| "n_total": 3107, |
| "groups_used": { |
| "VJepa2": 2048, |
| "DINOv3-B": 596, |
| "DINOv3-L": 209, |
| "DINOv2": 142, |
| "Multi-Backbone": 39, |
| "SigLIP2": 73 |
| } |
| }, |
| "encoder_anova": { |
| "f_statistic": 40.55315314313744, |
| "p_value": 1.1102230246251565e-16, |
| "df_between": 3, |
| "df_within": 3117, |
| "eta_squared": 0.037564758182648296, |
| "n_groups": 4, |
| "n_total": 3121, |
| "groups_used": { |
| "Zipformer": 2396, |
| "Retention": 337, |
| "Hybrid R-M": 95, |
| "BiMamba": 293 |
| } |
| }, |
| "backbone_x_encoder_anova": { |
| "f_statistic": 122.83218752862496, |
| "p_value": 1.1102230246251565e-16, |
| "df_between": 12, |
| "df_within": 3039, |
| "eta_squared": 0.3266099581471823, |
| "n_groups": 13, |
| "n_total": 3052, |
| "groups_used": { |
| "VJepa2+Zipformer": 1607, |
| "DINOv3-B+Zipformer": 413, |
| "DINOv3-B+Retention": 104, |
| "VJepa2+Retention": 170, |
| "DINOv3-L+Zipformer": 167, |
| "DINOv2+Zipformer": 130, |
| "VJepa2+Hybrid R-M": 68, |
| "VJepa2+BiMamba": 202, |
| "DINOv3-B+BiMamba": 72, |
| "Multi-Backbone+Zipformer": 24, |
| "DINOv3-L+Retention": 35, |
| "SigLIP2+Retention": 17, |
| "SigLIP2+Zipformer": 43 |
| } |
| }, |
| "backbone_stats": { |
| "convnext_small": { |
| "count": 2, |
| "mean": 0.9499983074246235, |
| "std": 0.015049833375802013, |
| "best": 0.9650481408004256 |
| }, |
| "eva02_large_patch14_448": { |
| "count": 4, |
| "mean": 0.8247990897522004, |
| "std": 0.020920762764346897, |
| "best": 0.8466378099729492 |
| }, |
| "hf:timm/fastvit_sa12.apple_in1k": { |
| "count": 1, |
| "mean": 0.8210893820116859, |
| "std": 0.0, |
| "best": 0.8210893820116859 |
| }, |
| "VJepa2": { |
| "count": 2048, |
| "mean": 0.8201971068945668, |
| "std": 0.24275298826263794, |
| "best": 1.0 |
| }, |
| "convnextv2_nano_fcmae_ft_in22k_in1k": { |
| "count": 3, |
| "mean": 0.8104509254098405, |
| "std": 0.07300397792978129, |
| "best": 0.9042845052901038 |
| }, |
| "DINOv2": { |
| "count": 142, |
| "mean": 0.7982659317729559, |
| "std": 0.16909593483906676, |
| "best": 0.9908533038136097 |
| }, |
| "DINOv3-L": { |
| "count": 209, |
| "mean": 0.7725610436169134, |
| "std": 0.17510321256700387, |
| "best": 0.9708420459837804 |
| }, |
| "InternViT": { |
| "count": 3, |
| "mean": 0.6742845235139282, |
| "std": 0.02803486309780819, |
| "best": 0.7038360494206984 |
| }, |
| "swin_large_patch4_window7_224": { |
| "count": 1, |
| "mean": 0.6258455518977457, |
| "std": 0.0, |
| "best": 0.6258455518977457 |
| }, |
| "SigLIP2": { |
| "count": 73, |
| "mean": 0.5338062135738759, |
| "std": 0.30061451723431637, |
| "best": 0.9957691426051499 |
| }, |
| "Multi-Backbone": { |
| "count": 39, |
| "mean": 0.5132637753316557, |
| "std": 0.09163961955875692, |
| "best": 0.7751574573174227 |
| }, |
| "DINOv3-B": { |
| "count": 596, |
| "mean": 0.3695597417612688, |
| "std": 0.3857176936458129, |
| "best": 0.99565846488945 |
| }, |
| "hf:apple/mobilevitv2-1.0-imagenet1k-256": { |
| "count": 1, |
| "mean": 0.10026244239376117, |
| "std": 0.0, |
| "best": 0.10026244239376117 |
| } |
| }, |
| "encoder_stats": { |
| "gru_temporal": { |
| "count": 1, |
| "mean": 0.9397300873785934, |
| "std": 0.0, |
| "best": 0.9397300873785934 |
| }, |
| "Zipformer": { |
| "count": 2396, |
| "mean": 0.752428300199627, |
| "std": 0.3088775095541608, |
| "best": 1.0 |
| }, |
| "Retention": { |
| "count": 337, |
| "mean": 0.6425876363598415, |
| "std": 0.33873447498561676, |
| "best": 0.990222269855006 |
| }, |
| "BiMamba": { |
| "count": 293, |
| "mean": 0.5849879690751034, |
| "std": 0.35285136869675116, |
| "best": 0.9935087343842928 |
| }, |
| "Hybrid R-M": { |
| "count": 95, |
| "mean": 0.5612062442661794, |
| "std": 0.3354911637185287, |
| "best": 0.9781107072499646 |
| } |
| } |
| }, |
| "analysis_5_agent_attribution": { |
| "agent_map_size": 5247, |
| "agent_stats": { |
| "Claude": { |
| "count": 321, |
| "mean_ap": 0.5642043867774518, |
| "std_ap": 0.3196754728101123, |
| "best_ap": 0.9852744878745394, |
| "median_ap": 0.7477192467371498, |
| "backbone_distribution": { |
| "VJepa2": 231, |
| "DINOv3-B": 50, |
| "DINOv3-L": 23, |
| "Multi-Backbone": 1, |
| "SigLIP2": 12, |
| "DINOv2": 4 |
| } |
| }, |
| "Gemini": { |
| "count": 192, |
| "mean_ap": 0.48348352650209997, |
| "std_ap": 0.35832911990476873, |
| "best_ap": 0.9643543795612995, |
| "median_ap": 0.6196625668838436, |
| "backbone_distribution": { |
| "VJepa2": 111, |
| "DINOv3-B": 58, |
| "Multi-Backbone": 2, |
| "DINOv3-L": 7, |
| "DINOv2": 5, |
| "SigLIP2": 9 |
| } |
| }, |
| "Unknown": { |
| "count": 9966, |
| "mean_ap": 0.8466969141530047, |
| "std_ap": 0.18081734838035712, |
| "best_ap": 1.0, |
| "median_ap": 0.8827026242765739, |
| "backbone_distribution": { |
| "VJepa2": 1706, |
| "DINOv3-B": 488, |
| "DINOv3-L": 179, |
| "DINOv2": 133, |
| "Multi-Backbone": 36, |
| "SigLIP2": 52, |
| "convnextv2_nano_fcmae_ft_in22k_in1k": 3, |
| "eva02_large_patch14_448": 4, |
| "InternViT": 3, |
| "hf:apple/mobilevitv2-1.0-imagenet1k-256": 1, |
| "swin_large_patch4_window7_224": 1, |
| "convnext_small": 2, |
| "hf:timm/fastvit_sa12.apple_in1k": 1 |
| } |
| } |
| } |
| }, |
| "analysis_6_nexar_competition": { |
| "competition_scores_found": 0, |
| "competition_details": [], |
| "leaderboard_info": { |
| "metric": "mAP", |
| "n_entries": 0 |
| } |
| }, |
| "generated_at": "2026-03-13T21:08:45.418086" |
| } |