{ "rule": "every (holdout, training) pair below the 1st percentile of the TRAINING set's own pairwise distance distribution is named, listened to, and if it is the same recording the holdout copy moves to training or both are excluded under E5", "training_pairs": 54615, "percentile": 1.0, "threshold_distance": 0.118752, "dataset_hash_source": "manifest.jsonl", "flagged_pairs": 44, "pairs": [ { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321718", "training_title": "Lust", "training_artist": "Oilboy's Aftersun", "distance": 0.058483, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 131.302, "holdout_duration_s": 269.991134, "training_duration_s": 401.293061 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321732", "training_title": "I wanna dance with you", "training_artist": "Oilboy's Aftersun", "distance": 0.068087, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 152.511, "holdout_duration_s": 327.096803, "training_duration_s": 479.608163 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_0295178", "training_title": "Wady-Funk Elek", "training_artist": "Adrian Sanchez DJ", "distance": 0.084507, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 86.57, "holdout_duration_s": 210.442449, "training_duration_s": 123.872653 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321735", "training_title": "Insane", "training_artist": "Oilboy's Aftersun", "distance": 0.085895, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 105.153, "holdout_duration_s": 269.991134, "training_duration_s": 375.14449 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321729", "training_title": "Bedroom rhythm", "training_artist": "Oilboy's Aftersun", "distance": 0.092648, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 153.269, "holdout_duration_s": 327.096803, "training_duration_s": 480.365714 } }, { "holdout_track": "track_0253540", "holdout_title": "20th century", "holdout_artist": "Windpearl", "training_track": "track_0012410", "training_title": "Feel You", "training_artist": "#Dance 75#", "distance": 0.092698, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 740.443, "holdout_duration_s": 905.164263, "training_duration_s": 164.721429 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321733", "training_title": "Summer slammer", "training_artist": "Oilboy's Aftersun", "distance": 0.093074, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 6.541, "holdout_duration_s": 269.991134, "training_duration_s": 276.532245 } }, { "holdout_track": "track_0253540", "holdout_title": "20th century", "holdout_artist": "Windpearl", "training_track": "track_1336432", "training_title": "We feel the same", "training_artist": "Index", "distance": 0.094573, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 600.183, "holdout_duration_s": 905.164263, "training_duration_s": 304.981338 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321741", "training_title": "When your world is getting crazy", "training_artist": "Oilboy's Aftersun", "distance": 0.095293, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 22.866, "holdout_duration_s": 327.096803, "training_duration_s": 349.962449 } }, { "holdout_track": "track_1395651", "holdout_title": "Deep Fall", "holdout_artist": "One Milk", "training_track": "track_1414779", "training_title": "Signal", "training_artist": "Terepin", "distance": 0.098319, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 19.06, "holdout_duration_s": 384.010408, "training_duration_s": 364.95 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_1321731", "training_title": "Oil boiled fantasies", "training_artist": "Oilboy's Aftersun", "distance": 0.099352, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 129.228, "holdout_duration_s": 210.442449, "training_duration_s": 339.670204 } }, { "holdout_track": "track_1086156", "holdout_title": "Genie Of Head (Radio Edit)", "holdout_artist": "Diavoloops", "training_track": "track_1068103", "training_title": "Get On The Floor", "training_artist": "GYAKO", "distance": 0.099385, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 17.713, "holdout_duration_s": 221.538231, "training_duration_s": 203.824762 } }, { "holdout_track": "track_1086156", "holdout_title": "Genie Of Head (Radio Edit)", "holdout_artist": "Diavoloops", "training_track": "track_1071079", "training_title": "You Need", "training_artist": "Luigi Talluto", "distance": 0.099568, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 41.421, "holdout_duration_s": 221.538231, "training_duration_s": 180.117506 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321716", "training_title": "Dont get me down", "training_artist": "Oilboy's Aftersun", "distance": 0.099587, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 135.037, "holdout_duration_s": 269.991134, "training_duration_s": 405.028571 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_1321724", "training_title": "Funkd up", "training_artist": "Oilboy's Aftersun", "distance": 0.100924, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 219.193, "holdout_duration_s": 210.442449, "training_duration_s": 429.635918 } }, { "holdout_track": "track_0253540", "holdout_title": "20th century", "holdout_artist": "Windpearl", "training_track": "track_1091136", "training_title": "House", "training_artist": "Index", "distance": 0.101583, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 614.54, "holdout_duration_s": 905.164263, "training_duration_s": 290.624649 } }, { "holdout_track": "track_0253672", "holdout_title": "Homo Sapiens Deletantis", "holdout_artist": "Windpearl", "training_track": "track_1091136", "training_title": "House", "training_artist": "Index", "distance": 0.104483, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 21.381, "holdout_duration_s": 312.005215, "training_duration_s": 290.624649 } }, { "holdout_track": "track_0704836", "holdout_title": "Snabisch Creator - Akiam Dance", "holdout_artist": "Snabisch", "training_track": "track_0012414", "training_title": "Neverever", "training_artist": "#Dance 75#", "distance": 0.104595, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 161.896, "holdout_duration_s": 294.911497, "training_duration_s": 133.01551 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_0431947", "training_title": "ADRIAN SANCHEZ WHATEVER GIRL - ACTIVATOR", "training_artist": "Adrian Sanchez DJ", "distance": 0.104858, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 92.902, "holdout_duration_s": 269.991134, "training_duration_s": 362.893061 } }, { "holdout_track": "track_1051920", "holdout_title": "Ibiza's twilight", "holdout_artist": "Dj Saryon & Sory", "training_track": "track_1121417", "training_title": "Meteor", "training_artist": "Marco Margna", "distance": 0.105354, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 135.504, "holdout_duration_s": 216.964989, "training_duration_s": 352.468594 } }, { "holdout_track": "track_0348704", "holdout_title": "THE LOVERS", "holdout_artist": "BRAJAN B&B", "training_track": "track_1297477", "training_title": "Mr.DeVo!-Go fuck yourself!", "training_artist": "Mr.DeVo!", "distance": 0.105665, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 60.537, "holdout_duration_s": 286.538209, "training_duration_s": 347.075283 } }, { "holdout_track": "track_1061536", "holdout_title": "Find Me (Orginal Mix)", "holdout_artist": "Alex Vinyl", "training_track": "track_1305122", "training_title": "Left behind", "training_artist": "Shopping Zombies", "distance": 0.106783, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 350.912, "holdout_duration_s": 91.590249, "training_duration_s": 442.502585 } }, { "holdout_track": "track_1086156", "holdout_title": "Genie Of Head (Radio Edit)", "holdout_artist": "Diavoloops", "training_track": "track_1168538", "training_title": "Cura Gyako 2014", "training_artist": "GYAKO", "distance": 0.107261, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 31.297, "holdout_duration_s": 221.538231, "training_duration_s": 190.241088 } }, { "holdout_track": "track_0921856", "holdout_title": "Forget the Pain (Instrumental Mix)", "holdout_artist": "Robert Chacón", "training_track": "track_1146085", "training_title": "Feel The Rhythm", "training_artist": "Synthron", "distance": 0.107635, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 206.492, "holdout_duration_s": 281.25, "training_duration_s": 487.74195 } }, { "holdout_track": "track_0704836", "holdout_title": "Snabisch Creator - Akiam Dance", "holdout_artist": "Snabisch", "training_track": "track_1121446", "training_title": "Freedom", "training_artist": "Marco Margna", "distance": 0.108576, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 24.341, "holdout_duration_s": 294.911497, "training_duration_s": 319.252766 } }, { "holdout_track": "track_1086156", "holdout_title": "Genie Of Head (Radio Edit)", "holdout_artist": "Diavoloops", "training_track": "track_1196501", "training_title": "DJ DeadSheakerz feat. HunterSynth - Preoccupied in Thoughts (Craig Ferrier Remix)", "training_artist": "deadsheakerz", "distance": 0.108627, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 119.676, "holdout_duration_s": 221.538231, "training_duration_s": 341.214286 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321726", "training_title": "You read me wrong", "training_artist": "Oilboy's Aftersun", "distance": 0.109302, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 68.371, "holdout_duration_s": 327.096803, "training_duration_s": 395.467755 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321731", "training_title": "Oil boiled fantasies", "training_artist": "Oilboy's Aftersun", "distance": 0.11004, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 69.679, "holdout_duration_s": 269.991134, "training_duration_s": 339.670204 } }, { "holdout_track": "track_1290255", "holdout_title": "Arppleggiator (Long Version)", "holdout_artist": "Apple Victims", "training_track": "track_1414796", "training_title": "Signal Redux", "training_artist": "Terepin", "distance": 0.111073, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 53.081, "holdout_duration_s": 304.41941, "training_duration_s": 357.5 } }, { "holdout_track": "track_1129545", "holdout_title": "I Just Want..", "holdout_artist": "Diavoloops", "training_track": "track_0200574", "training_title": "genius", "training_artist": "bernimusic", "distance": 0.111944, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 6.0, "holdout_duration_s": 294.0, "training_duration_s": 288.0 } }, { "holdout_track": "track_1051920", "holdout_title": "Ibiza's twilight", "holdout_artist": "Dj Saryon & Sory", "training_track": "track_1321732", "training_title": "I wanna dance with you", "training_artist": "Oilboy's Aftersun", "distance": 0.112983, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 262.643, "holdout_duration_s": 216.964989, "training_duration_s": 479.608163 } }, { "holdout_track": "track_1246352", "holdout_title": "Solar Breeze", "holdout_artist": "Elenesski", "training_track": "track_1104259", "training_title": "Dj_Quincy_Ortiz_-_Groove_Soul", "training_artist": "Dj Quincy Ortiz", "distance": 0.112994, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 198.732, "holdout_duration_s": 180.983628, "training_duration_s": 379.715918 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321723", "training_title": "Natural instinct", "training_artist": "Oilboy's Aftersun", "distance": 0.113054, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 407.495, "holdout_duration_s": 269.991134, "training_duration_s": 677.485714 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1308623", "training_title": "Nightshift", "training_artist": "ZZZZZ", "distance": 0.114399, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 498.968, "holdout_duration_s": 269.991134, "training_duration_s": 768.958821 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_1321718", "training_title": "Lust", "training_artist": "Oilboy's Aftersun", "distance": 0.115469, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 190.851, "holdout_duration_s": 210.442449, "training_duration_s": 401.293061 } }, { "holdout_track": "track_0704836", "holdout_title": "Snabisch Creator - Akiam Dance", "holdout_artist": "Snabisch", "training_track": "track_1414795", "training_title": "Technology Redux", "training_artist": "Terepin", "distance": 0.115896, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 16.389, "holdout_duration_s": 294.911497, "training_duration_s": 311.3 } }, { "holdout_track": "track_1036656", "holdout_title": "Adore - feat. Linde Sagen", "holdout_artist": "Sweden House Project", "training_track": "track_1312908", "training_title": "the house-luigi talluto", "training_artist": "Luigi Talluto", "distance": 0.116028, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 27.272, "holdout_duration_s": 206.272426, "training_duration_s": 179.0 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321735", "training_title": "Insane", "training_artist": "Oilboy's Aftersun", "distance": 0.116472, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 48.048, "holdout_duration_s": 327.096803, "training_duration_s": 375.14449 } }, { "holdout_track": "track_1290255", "holdout_title": "Arppleggiator (Long Version)", "holdout_artist": "Apple Victims", "training_track": "track_1414795", "training_title": "Technology Redux", "training_artist": "Terepin", "distance": 0.116649, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 6.881, "holdout_duration_s": 304.41941, "training_duration_s": 311.3 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321717", "training_title": "Together in harmony", "training_artist": "Oilboy's Aftersun", "distance": 0.116824, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 52.961, "holdout_duration_s": 269.991134, "training_duration_s": 322.951837 } }, { "holdout_track": "track_1154952", "holdout_title": "A.K.1974 - [Call Me] If You Need Me", "holdout_artist": "A.K.1974", "training_track": "track_1321714", "training_title": "Naked, but still too hot to handle", "training_artist": "Oilboy's Aftersun", "distance": 0.117365, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 73.441, "holdout_duration_s": 269.991134, "training_duration_s": 343.431837 } }, { "holdout_track": "track_0421639", "holdout_title": "Skinny Model (thin is in remix)", "holdout_artist": "Emorej", "training_track": "track_1321716", "training_title": "Dont get me down", "training_artist": "Oilboy's Aftersun", "distance": 0.117995, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 77.932, "holdout_duration_s": 327.096803, "training_duration_s": 405.028571 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_1321716", "training_title": "Dont get me down", "training_artist": "Oilboy's Aftersun", "distance": 0.118205, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 194.586, "holdout_duration_s": 210.442449, "training_duration_s": 405.028571 } }, { "holdout_track": "track_0212013", "holdout_title": "yepa", "holdout_artist": "Tet production", "training_track": "track_1321714", "training_title": "Naked, but still too hot to handle", "training_artist": "Oilboy's Aftersun", "distance": 0.118436, "objective_screen": { "normalised_title_match": false, "same_artist": false, "byte_identical_audio": false, "duration_delta_s": 132.989, "holdout_duration_s": 210.442449, "training_duration_s": 343.431837 } } ], "consequence": "absent this scan the split is an ARTIST split, not a WORK split; with it, every flagged pair is named here", "run_id": "20260906T092541Z-house0", "generated_utc": "2026-09-06T09:35:26Z", "status": "RUN — before any preprocessing, on the A1-pinned CLAP embeddings", "objective_screen_summary": { "flagged_pairs": 44, "distinct_holdout_tracks": 16, "distinct_training_tracks": 35, "normalised_title_matches": 0, "same_artist_pairs": 0, "byte_identical_audio": 0, "duration_within_1s": 0, "distance_range": [ 0.058483, 0.118436 ], "neighbours_per_flagged_holdout_track": { "track_1154952": 10, "track_0421639": 6, "track_0212013": 6, "track_1086156": 4, "track_0253540": 3, "track_0704836": 3, "track_1051920": 2, "track_1290255": 2, "track_1395651": 1, "track_0253672": 1, "track_0348704": 1, "track_1061536": 1, "track_0921856": 1, "track_1129545": 1, "track_1246352": 1, "track_1036656": 1 } }, "reading": "NO PAIR CARRIES A DUPLICATE SIGNATURE. Zero of the 44 share a normalised title, zero share an artist (impossible under an artist split, and computed rather than assumed), zero share audio bytes, and zero are within 1.0 s of each other's duration. The shape is one-to-many, not one-to-one: 44 pairs come from just 16 holdout tracks, one of which is near TEN training tracks. A re-upload of the same recording is near exactly one thing; a track sitting in the dense middle of a one-genre corpus is near many. So the registered action -- 'if it is the same recording the holdout copy moves to the training side or both are excluded under E5' -- DOES NOT FIRE on the objective evidence, and no track moved: the split is byte-identical to the one frozen at 09:25:41Z.", "still_owed": "The rule also says the flagged pairs are LISTENED TO. That is the operator's ear and it has not happened. Until it does, this scan is an objective screen that found no duplicate signature, not a confirmed work split, and the write-up says exactly that.", "why_this_is_not_inside_split_json": "PREREG 3.1(7) says flagged pairs are 'named in split.json'. split.json was already sha-pinned into freeze_point (dataset tier) and dataset/MANIFEST.sha256 when this scan completed, and the scan moved no track. Rewriting it would have silently changed the freeze identity to record a measurement that changed nothing. The scan is therefore filed here, beside the freeze, and split.json's own field still reads 'PENDING' because that is what was true AT THE FREEZE. Named as a deviation." }