{
 "schema_version": "1.0",
 "page_kind": "article",
 "pack_kind": "generate",
 "slug": "ten-minutes-with-living-artists",
 "title": "Ten Minutes with Living Artists",
 "dek": "A genre instead of a composer: 159 tracks of Creative Commons minimal techno, the same ten-minute recipe that taught three dead composers, and the one ear in the house preferred the model with nothing added — both times.",
 "published": "2026-09-03",
 "series": [
  "guide",
  "bench"
 ],
 "licence": "CC BY 4.0",
 "status": "pending-judge",
 "pack_note": "no judge is seated, so every chip is held",
 "notice": "items with published:false are held and are not this page's published words",
 "source": {
  "url": "https://research.strata2signal.com/ten-minutes-with-living-artists/",
  "md_url": "https://research.strata2signal.com/ten-minutes-with-living-artists/index.md",
  "md_sha": "b37b5bf7cecfbd1dc63df375b05f522958abae98a3fc7a2b655e42ef593954a6",
  "html_sha": "8f2506ea5dbc26b0473cba55efa56fc764210f59b4d3e28ad598aa6affd1046b",
  "anchors_sha": "926fe2c3e6816da113a6cc1fcc2ee9ab5697caaf603e0f139c0e5b843d7e4d7e"
 },
 "short": {
  "paragraph": "A genre instead of a composer: 159 tracks of Creative Commons minimal techno, the same ten-minute recipe that taught three dead composers, and the one ear in the house preferred the model with nothing added - both times.",
  "from_dek": true,
  "counts": {
   "words": 6958,
   "minutes": 32,
   "tables": 7,
   "kit": true
  },
  "bullets": [
   {
    "text": "the training error reached its lowest point at pass nine with a value of 0.633",
    "figure": "0.633",
    "cite": "ten-minutes-on-the-big-card",
    "quote": "The per-pass training error fell from 0.99 to 0.65, but not smoothly: down through pass seven, up at eight, down to the run's lowest at nine - 0.633 - and up again at ten.",
    "span": [
     14215,
     14223
    ]
   }
  ]
 },
 "sections": [
  {
   "id": "where-part-two-left-off",
   "heading": "Where part two left off",
   "level": 2,
   "span": [
    1256,
    1954
   ],
   "chunks": [
    [
     1256,
     1954
    ]
   ],
   "chars": 698,
   "digest": "the article recaps the machinery used in previous parts, including an open-weights music generator, a detachable adapter trained with LoRA, a strength dial, and a name tag. it notes that the fourth adapter's best training pass was never saved to disk.",
   "digest_skipped": null
  },
  {
   "id": "a-genre-not-a-composer",
   "heading": "A genre, not a composer",
   "level": 2,
   "span": [
    1954,
    3408
   ],
   "chunks": [
    [
     1954,
     3408
    ]
   ],
   "chars": 1454,
   "digest": null,
   "digest_skipped": null
  },
  {
   "id": "the-ear",
   "heading": "The ear",
   "level": 2,
   "span": [
    3408,
    6787
   ],
   "chunks": [
    [
     3408,
     6787
    ]
   ],
   "chars": 3379,
   "digest": "a human operator performed unblinded listening on 2026-08-31. the process used specific requests, seeds, and normalization at −1 dB. the operator concluded that the minimal techno adapter did not perform well, suggesting the model itself might be the weak instrument.",
   "digest_skipped": null
  },
  {
   "id": "forty-one-releases-licence-by-licence",
   "heading": "Forty-one releases, licence by licence",
   "level": 2,
   "span": [
    6787,
    10356
   ],
   "chunks": [
    [
     6787,
     10356
    ]
   ],
   "chars": 3569,
   "digest": null,
   "digest_skipped": "credits"
  },
  {
   "id": "the-question-in-court",
   "heading": "The question in court",
   "level": 2,
   "span": [
    10356,
    12437
   ],
   "chunks": [
    [
     10356,
     12437
    ]
   ],
   "chars": 2081,
   "digest": "the text mentions a lawsuit by Sony Music Publishing and Warner Chappell against Anthropic regarding training on copyrighted works. the authors state they use only music given away by dedication, age, or licence, and they provide full attribution for the adapter.",
   "digest_skipped": null
  },
  {
   "id": "ten-minutes-on-the-big-card",
   "heading": "Ten minutes on the big card",
   "level": 2,
   "span": [
    12437,
    15740
   ],
   "chunks": [
    [
     12437,
     15740
    ]
   ],
   "chars": 3303,
   "digest": "training took 10 minutes 8 seconds on a workstation-class card. the process involved 400 updates. the shipped checkpoint was pass ten, though pass nine had a lower error of 0.633. the adapter contains 44 million adjustable weights and is 88,130,280 bytes.",
   "digest_skipped": null
  },
  {
   "id": "was-it-the-adapter-or-the-words",
   "heading": "Was it the adapter, or the words?",
   "level": 2,
   "span": [
    15740,
    18103
   ],
   "chunks": [
    [
     15740,
     18103
    ]
   ],
   "chars": 2363,
   "digest": "testing showed that the shape of the caption affects loudness. on corpus-shaped requests, the adapter moved loudness by −0.22 and +0.49. the round demonstrated that the caption shape is a variable that must be pinned before blaming the adapter.",
   "digest_skipped": null
  },
  {
   "id": "one-artist-instead-of-a-tag",
   "heading": "One artist instead of a tag",
   "level": 2,
   "span": [
    18103,
    25524
   ],
   "chunks": [
    [
     18103,
     25524
    ]
   ],
   "chars": 7421,
   "digest": "a control using one artist, Floating Mind, was trained for 1 minute 33 seconds. the coherence bench showed the tag-assembled corpus has a spread of 0.4489, which is about 1.9 times further apart than the one-artist corpus.",
   "digest_skipped": null
  },
  {
   "id": "twelve-requests-four-ways",
   "heading": "Twelve requests, four ways",
   "level": 2,
   "span": [
    25524,
    30225
   ],
   "chunks": [
    [
     25524,
     30225
    ]
   ],
   "chars": 4701,
   "digest": "twelve requests were rendered four ways to test the impact of the adapter and genre. the results showed that the adapter's effect on loudness varies by request, and the model's own ceiling can cause flat-topped stretches in the audio.",
   "digest_skipped": null
  },
  {
   "id": "what-we-did-not-measure",
   "heading": "What we did NOT measure",
   "level": 2,
   "span": [
    30225,
    31472
   ],
   "chunks": [
    [
     30225,
     31472
    ]
   ],
   "chars": 1247,
   "digest": "the authors list several unmeasured factors, including the quality of monitors, whether the model is good at the genre, and whether a DJ-curated subset would yield different results. they note that the coherence bench cannot calibrate against an ear.",
   "digest_skipped": null
  },
  {
   "id": "what-to-take-with-you",
   "heading": "What to take with you",
   "level": 2,
   "span": [
    31472,
    33053
   ],
   "chunks": [
    [
     31472,
     33053
    ]
   ],
   "chars": 1581,
   "digest": "the article concludes that the corpus was the primary difference in results. it emphasizes that the caption is a variable, the best pass was not saved, and the model itself is the weak instrument for making good edm.",
   "digest_skipped": null
  },
  {
   "id": "how-to-check-our-work",
   "heading": "How to check our work",
   "level": 2,
   "span": [
    33053,
    34341
   ],
   "chunks": [
    [
     33053,
     34341
    ]
   ],
   "chars": 1288,
   "digest": null,
   "digest_skipped": "credits"
  },
  {
   "id": "who-ran-this-and-thanks",
   "heading": "Who ran this, and thanks",
   "level": 2,
   "span": [
    34341,
    45244
   ],
   "chunks": [
    [
     34341,
     45244
    ]
   ],
   "chars": 10903,
   "digest": null,
   "digest_skipped": "credits"
  },
  {
   "id": "the-rest-of-the-seminar",
   "heading": "The rest of the seminar",
   "level": 2,
   "span": [
    45244,
    45833
   ],
   "chunks": [
    [
     45244,
     45833
    ]
   ],
   "chars": 589,
   "digest": null,
   "digest_skipped": "credits"
  }
 ],
 "chips": [
  {
   "id": "c-bd725abe",
   "q": "how does the music generation machinery function?",
   "a": "The process uses an open-weights music generator called ACE-Step 1.5. It employs a small detachable adapter trained with LoRA, a strength dial ranging from zero to full, and a name tag consisting of a made-up first word in the request to identify the specific adapter.",
   "cites": [
    "where-part-two-left-off"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-dca7177c",
   "q": "how were the audio clips prepared for listening?",
   "a": "Clips were rendered with specific parameters: 128 beats per minute, A minor, four-four time, and instrumental. The engine used output normalisation at −1 dB to cap peaks at the same ceiling, ensuring differences resulted from the adapter rather than volume variations.",
   "cites": [
    "the-ear"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-656ee86c",
   "q": "what legal context surrounds the training of these models?",
   "a": "The article mentions a lawsuit by Sony Music Publishing and Warner Chappell against Anthropic, alleging illegal scraping of copyrighted works. The authors note that the adapters in this study use music that was provided in advance via dedication, age, or specific licences.",
   "cites": [
    "the-question-in-court"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-65423936",
   "q": "how long did the training process take?",
   "a": "The training involved ten passes over 159 tracks, totaling 400 weight updates. The ten passes themselves took 10 minutes 6 seconds, resulting in an average of 1.515 seconds per update.",
   "cites": [
    "ten-minutes-on-the-big-card"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-ef588f67",
   "q": "how does caption shape affect the audio output?",
   "a": "Using corpus-shaped requests, such as 'minimal techno, title', resulted in different loudness shifts compared to descriptive prose requests. While descriptive requests caused the adapter to make audio quieter, corpus-shaped requests moved loudness by −0.22 and +0.49 LUFS.",
   "cites": [
    "was-it-the-adapter-or-the-words"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-9d38d640",
   "q": "what was the result of the coherence bench test?",
   "a": "The bench measured the spread of corpora using CLAP. The tag-assembled minimal techno corpus had a spread of 0.4489, which was about 1.9 times further apart than the one-artist control corpus, which had a spread of 0.2378.",
   "cites": [
    "one-artist-instead-of-a-tag"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  },
  {
   "id": "c-fdcd41f0",
   "q": "what does the loudness data reveal about the twelve requests?",
   "a": "Across twelve requests, the meter moved about twice as far on the six techno requests nearest the corpus compared to the six farther away. Specifically, the techno requests averaged 1.35 LUFS of movement, while the others averaged 0.66 LUFS.",
   "cites": [
    "twelve-requests-four-ways"
   ],
   "published": false,
   "publish_state": "held",
   "quote": "",
   "quote_cite": ""
  }
 ],
 "related": [
  {
   "slug": "the-ceiling-is-not-the-corpus",
   "why": "shares ground with § Twelve requests, four ways · § What the adapter actually did"
  },
  {
   "slug": "listen-for-yourself",
   "why": "shares ground with § Twelve requests, four ways · § What to listen for"
  },
  {
   "slug": "half-an-hour-with-dead-composers",
   "why": "shares ground with § Where part two left off · § Previously, in five minutes"
  }
 ],
 "thanks": "## Who ran this, and thanks {#who-ran-this-and-thanks}\n\nThe corpus first, because this time it is owed. Forty-one releases from the Internet Archive's\nnetlabel collections under CC0, CC BY and CC BY-SA, listed in full after this section: each row links the archive item it came from and names the licence recorded on that item, and the licence instruments are spelled out, deed by deed, above the tables. Two rows read \"creator not stated\nin the record,\" because an invented artist name in a credits list would be the worst possible\nbug. One netlabel's name contains an expletive; it is the legally correct attribution string\nand it is printed as recorded. The model: ACE-Step 1.5 (MIT — its row on our\n[licence ledger](/licences/)); its model card asks that AI involvement be disclosed, and every\nclip on this page is AI-generated and says so. The trainer: [Side-Step](https://github.com/koda-dernet/Side-Step), the corrected-timestep LoRA trainer vendored in ACE-Step\n1.5; its upstream licence is CC BY-NC-SA 4.0, though the vendored copy says it follows\nACE-Step's MIT — we proceed on the stricter reading, and wrote to ask. The coherence\ninstrument: LAION's CLAP — the `laion_clap` package at 1.1.7 and the `lukewys/laion_clap`\ncheckpoint host that carries the music weights the bench ran on; the licence file on disk\nreads CC0 1.0 for the code and the checkpoint repository declares the same, though the\npackage's own index metadata says Apache 2.0 — both permissive, nothing turns on it.\n\nAnd beneath all of it, the open tools this work stood on without modifying: PyTorch and\ntorchaudio, Hugging Face's PEFT library, NumPy, FFmpeg — which decoded every training file, cut\nevery crop and measured every loudness figure — and the Internet Archive, which carried all\nforty-one releases and every licence record we read, fetched under a user agent that names us\nand says how to reach us: `strata2signal-research/1.0 (private research; contact via\nstrata2signal.com)`. None of them owed us anything. A small team and a fleet of AI agents did\nthe work; the humans signed the numbers — [how that works](/how-we-work/) is next door.\n\n**The licence instruments.** Each label in the tables stands for the deed at the address\nbeside it; the items record each address with an http or https scheme, and the dedication\nappears under both, which is why ten spellings cover nine deeds.\n\n| label | deed |\n|---|---|\n| CC0 | [creativecommons.org/publicdomain/zero/1.0/](https://creativecommons.org/publicdomain/zero/1.0/) |\n| BY 1.0 | [creativecommons.org/licenses/by/1.0/](https://creativecommons.org/licenses/by/1.0/) |\n| BY 3.0 | [creativecommons.org/licenses/by/3.0/](https://creativecommons.org/licenses/by/3.0/) |\n| BY 4.0 | [creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/) |\n| BY-SA 2.5 | [creativecommons.org/licenses/by-sa/2.5/](https://creativecommons.org/licenses/by-sa/2.5/) |\n| BY-SA 3.0 | [creativecommons.org/licenses/by-sa/3.0/](https://creativecommons.org/licenses/by-sa/3.0/) |\n| BY-SA 3.0 CH | [creativecommons.org/licenses/by-sa/3.0/ch/](https://creativecommons.org/licenses/by-sa/3.0/ch/) |\n| BY-SA 3.0 US | [creativecommons.org/licenses/by-sa/3.0/us/](https://creativecommons.org/licenses/by-sa/3.0/us/) |\n| BY-SA 4.0 | [creativecommons.org/licenses/by-sa/4.0/](https://creativecommons.org/licenses/by-sa/4.0/) |\n\nEach release's identifier links to its archive item.\n\n**CC0 — nine releases, nothing owed, credited anyway.**\n\n| release | creator | files | archive item | notes |\n|---|---|---|---|---|\n| RW-Techordings presents [RWT-012] 30-hcir & Bertha James - Split One | Richard Wilmer | 2 | [30-hcir-berthaJamesSplit-RWT-012](https://archive.org/details/30-hcir-berthaJamesSplit-RWT-012) | in the corpus |\n| Black Saturday | ArchivOne | 1 | [archivone-black-saturday](https://archive.org/details/archivone-black-saturday) | in the corpus |\n| Flourish // Perish | Braids | 10 | [braids-flurish-perish](https://archive.org/details/braids-flurish-perish) | in the corpus |\n| RW-Techordings presents [RWT-010] ElectRICHual - Predictable (Mixes) | ElectRICHual | 2 | [ElectRICHual-Predictable-Mixes](https://archive.org/details/ElectRICHual-Predictable-Mixes) | in the corpus |\n| Hardware Techno | David Murillo Diaz | 24 | [hardware-techno](https://archive.org/details/hardware-techno) | in the corpus |\n| In Novation | Kλпξiðλ | 1 | [in-novation](https://archive.org/details/in-novation) | in the corpus |\n| knolios_moments | knolios | 6 | [knolios_moments](https://archive.org/details/knolios_moments) | in the corpus |\n| Mr.Dee_D1 | Mr.Dee | 3 | [Mr.Dee_D1](https://archive.org/details/Mr.Dee_D1) | in the corpus |\n| Dawn | Shinji Wakasa | 12 | [shinji-wakasa-dawn](https://archive.org/details/shinji-wakasa-dawn) | in the corpus |\n\n**CC BY — nine releases, attribution required.**\n\n| release | creator | licence | files | archive item | notes |\n|---|---|---|---|---|---|\n| Aaron Goldbody Vs. Luke The Wizard - Z Veseljem Exported EP [kahvi019] | Aaron goldbody vs. Luke the wizard | BY 1.0 | 2 | [kahvi019](https://archive.org/details/kahvi019) | in the corpus |\n| VA-Let the bass ruin your speakers [ONMP215] | VA | BY 3.0 | 24 | [onmp215a](https://archive.org/details/onmp215a) | in the corpus |\n| [shoki005g] - Various - Grey EP | Shoki Recordings | BY 3.0 | 4 | [shoki005g](https://archive.org/details/shoki005g) | in the corpus |\n| STE34. Jyolrstion (kind of placeless interlude) | krai | BY 3.0 | 5 | [STE34](https://archive.org/details/STE34) | in the corpus |\n| Nisiru Remixed EP | stroboskop | BY 3.0 | 6 | [stroboskop-label013](https://archive.org/details/stroboskop-label013) | in the corpus |\n| Reho Remixed | stroboskop | BY 4.0 | 6 | [Stroboskop033](https://archive.org/details/Stroboskop033) | in the corpus |\n| [Tranz023] Holocaos- Metamorfose Computador EP | Caue Miranda | BY 4.0 | 8 | [tranz023Holocaos-MetamorfoseComputadorEp_140](https://archive.org/details/tranz023Holocaos-MetamorfoseComputadorEp_140) | in the corpus |\n| [unfound88] happy in novi sad | creator not stated in the record | BY 4.0 | 15 | [unfound88](https://archive.org/details/unfound88) | in the corpus |\n| [unfound91] jukka-pekka kervinen - coffee beans | jukka-pekka kervinen | BY 4.0 | 4 | [unfound91](https://archive.org/details/unfound91) | in the corpus |\n\n**CC BY-SA — twenty-three releases, attribution required and share-alike.** Ten are the\none-artist control's corpus; two were kept on disk and excluded from training for their 16 kHz\nsource; the notes column says which.\n\n| release | creator | licence | files | archive item | notes |\n|---|---|---|---|---|---|\n| Chuänchö - Profilaxis [inoquo001] | Chuänchö | BY-SA 2.5 | 4 | [inoquo001](https://archive.org/details/inoquo001) | in the corpus |\n| [inoquo018] Ol - liturgy | Ol | BY-SA 2.5 | 4 | [inoquo018](https://archive.org/details/inoquo018) | in the corpus |\n| Tomorrow EP | Derek Scott | BY-SA 2.5 | 3 | [RAR005_Derek_Scott_Tomorrow_EP](https://archive.org/details/RAR005_Derek_Scott_Tomorrow_EP) | in the corpus |\n| Sequentialwork - Sequentialwork EP [bump188] | Kiyoshi Tomehara | BY-SA 3.0 | 3 | [bump188](https://archive.org/details/bump188) | in the corpus |\n| [inoQuo070] v.a. - we are alive | inoQuo | BY-SA 3.0 | 7 | [inoQuo070](https://archive.org/details/inoQuo070) | in the corpus |\n| [monoKraK 134] Sin Amigos \"Ningun Amigo\" | Sin Amigos | BY-SA 3.0 CH | 2 | [monokrak134SinAmigosnignAmigo](https://archive.org/details/monokrak134SinAmigosnignAmigo) | in the corpus |\n| (monoKraK173) Yann Detroit VS Floating Mind \"Dust\" | creator not stated in the record | BY-SA 3.0 | 2 | [MonoKraK173YannDetroitVSFloatingMind_Dust](https://archive.org/details/MonoKraK173YannDetroitVSFloatingMind_Dust) | in the corpus |\n| (monoKraK197) Albert Negredo \"Lithium Serendipity\" | Albert Negredo | BY-SA 3.0 | 2 | [MonoKraK197AlbertNegredoLithiumSerendipity](https://archive.org/details/MonoKraK197AlbertNegredoLithiumSerendipity) | in the corpus |\n| [monokrak199] Floating Mind \"Schöni\" | Floating Mind | BY-SA 3.0 | 3 | [Monokrak199FloatingMind_Schni](https://archive.org/details/Monokrak199FloatingMind_Schni) | control corpus |\n| [monoKraK200] Floating Mind \"Birthday Accelerated\" | Floating Mind | BY-SA 3.0 | 3 | [Monokrak200FloatingMind_Birthday_Accelerate](https://archive.org/details/Monokrak200FloatingMind_Birthday_Accelerate) | control corpus |\n| [monokrak 203] Floating Mind \"A Mind Is Floating\" | Floating Mind | BY-SA 3.0 | 3 | [Monokrak203FloatingMind_AMindIsFloating](https://archive.org/details/Monokrak203FloatingMind_AMindIsFloating) | control corpus |\n| [monoKraK 204] Floating Mind \"Disko Chill\" | Floating Mind | BY-SA 3.0 | 3 | [monoKraK204FloatingMind_DiskoChill](https://archive.org/details/monoKraK204FloatingMind_DiskoChill) | control corpus |\n| (monoKraK205) Floating Mind \"Voyager EP\" | Floating Mind | BY-SA 3.0 | 3 | [Monokrak205FloatingMind_VoyagerEP](https://archive.org/details/Monokrak205FloatingMind_VoyagerEP) | control corpus |\n| [monoKraK208] Floating Mind \"et si ...\" | Floating Mind | BY-SA 3.0 | 3 | [MonoKraK208FloatingMind_et_si](https://archive.org/details/MonoKraK208FloatingMind_et_si) | control corpus |\n| [monoKraK84] Various Artists \"Smoked mono vol.7\" | Zaid Edghaim, Kimo-S, Alicia Hush, Floating Mind | BY-SA 3.0 | 3 | [monokrak84VariousArtistssmokedMonoVol.7](https://archive.org/details/monokrak84VariousArtistssmokedMonoVol.7) | in the corpus |\n| [TFN110] Lik-o - TRASHFUCK NET EP | Lik-o | BY-SA 3.0 US | 3 | [TFN110](https://archive.org/details/TFN110) | in the corpus |\n| [TFN211] Graffiti Mechanism - Park | Graffiti Mechanism | BY-SA 3.0 US | 2 | [TFN211](https://archive.org/details/TFN211) | in the corpus |\n| [TFN223] Graffiti Mechanism - RENukeFamRMXS-EP | Graffiti Mechanism | BY-SA 3.0 US | 4 | [TFN223](https://archive.org/details/TFN223) | kept on disk, excluded from training (16 kHz source) |\n| [TFN247] Graffiti Mechanism - RE-RMXD-RENukeFamRMXS-EP | Graffiti Mechanism | BY-SA 3.0 US | 4 | [TFN247](https://archive.org/details/TFN247) | kept on disk, excluded from training (16 kHz source) |\n| [monokrak 209] Floating Mind \"Ready For Flying\" | Floating Mind | BY-SA 4.0 | 3 | [Monokrak209FloatingMind_Ready_For_Flying](https://archive.org/details/Monokrak209FloatingMind_Ready_For_Flying) | control corpus |\n| [monoKraK214] Floating Mind \"Hidden Passion\" | Floating Mind | BY-SA 4.0 | 3 | [Monokrak214FloatingMind_Hidden_Passion](https://archive.org/details/Monokrak214FloatingMind_Hidden_Passion) | control corpus |\n| [monokrak 216] Floating Mind \"Thru Lines\" | Floating Mind | BY-SA 4.0 | 3 | [Monokrak216FloatingMind_Thru_Lines](https://archive.org/details/Monokrak216FloatingMind_Thru_Lines) | control corpus |\n| [monokrak 217] Floating Mind \"Spatial Moments\" | Floating Mind | BY-SA 4.0 | 3 | [Monokrak217FloatingMind_Spatial_Moments](https://archive.org/details/Monokrak217FloatingMind_Spatial_Moments) | control corpus |",
 "kit": {
  "url": "https://research.strata2signal.com/ten-minutes-with-living-artists/data/",
  "licence": "CC BY 4.0",
  "files": [
   "PREREG-COHERENCE.md",
   "PREREG-ROUND4.md",
   "README.md",
   "bandwidth-audit.json",
   "clip-peaks-raw.tsv",
   "clips.json",
   "coherence/RESULTS.json",
   "coherence/RESULTS.md",
   "coherence/RESULTS.sha256",
   "coherence/RUN.log",
   "coherence/crop_windows.sh",
   "coherence/crops.csv",
   "coherence/embeddings-index.csv",
   "coherence/embeddings.npz",
   "coherence/metric1-corpus-spread.csv",
   "coherence/metric1-draws.csv",
   "coherence/metric2-pairs.csv",
   "coherence/metric2-render-centroid.csv",
   "coherence/run_bench.py",
   "corpus-manifest.json",
   "corpus-manifest.sha256",
   "corpus-provenance.md",
   "credits.md",
   "cross-arm-check-round1.json",
   "cross-arm-check-round2.json",
   "exclusions.json",
   "freeze-receipt.json",
   "intake-receipt.md",
   "records-round1.json",
   "records-round2.json",
   "records-round3.json",
   "records-round4.json",
   "round4-blind-sheet.json",
   "round4-captions-as-rendered.txt",
   "s-per-step.md",
   "split-fmctl0.json",
   "split-mnml0.json",
   "split-rule.py",
   "train-fmctl0.log",
   "train-mnml0.log",
   "unload-restore-control.json",
   "MANIFEST.sha256"
  ]
 },
 "seat_class": {
  "writer": "gemma-class",
  "judge": null,
  "writer_runtime": "vllm"
 },
 "bench": {
  "bullets_written": 3,
  "bullets_kept": 1,
  "digests_written": 10,
  "digests_kept": 9,
  "chips_written": 8,
  "chips_kept": 7,
  "chips_grounded": 0,
  "chips_published": 0,
  "dropped_by": {
   "new_noun": 2,
   "figure": 2,
   "length": 0,
   "cite": 0,
   "judge": 0,
   "quote": 0,
   "directive": 0,
   "redaction": 0,
   "profanity": 0
  },
  "new_noun_tokens_checked": [
   "0.22",
   "0.2378",
   "0.4489",
   "0.49",
   "0.633",
   "0.66",
   "1",
   "1.35",
   "1.5",
   "1.515",
   "1.9",
   "10",
   "128",
   "130",
   "159",
   "2026-08-31",
   "208",
   "280",
   "33",
   "400",
   "41",
   "44",
   "6",
   "63",
   "8",
   "88",
   "ACE-Step",
   "Anthropic",
   "Archive",
   "CLAP",
   "Chappell",
   "DJ",
   "DJ-curated",
   "Floating",
   "Internet",
   "LUFS",
   "LoRA",
   "Mind",
   "Music",
   "Publishing",
   "Sony",
   "Warner"
  ],
  "new_noun_tokens_withheld": 2,
  "figure_definition": "v3",
  "drops": [
   {
    "kind": "digest",
    "reason": "figure",
    "item": "the focus shifts from composers to the genre of minimal techno. the training corpus consisted of 208 tracks from 41 releases found on the Internet Archive. the DJ excluded 63 files that ran twenty min",
    "detail": "figures absent: ['41']"
   },
   {
    "kind": "bullet",
    "reason": "figure",
    "item": "the training for the wide run included 159",
    "detail": "terminal token not verbatim in the cited span"
   },
   {
    "kind": "bullet",
    "reason": "new_noun",
    "item": null,
    "item_chars": 4,
    "withheld": "a string naming something the article does not",
    "detail": ""
   },
   {
    "kind": "chip",
    "reason": "new_noun",
    "item": null,
    "item_chars": 4,
    "withheld": "a string naming something the article does not",
    "detail": ""
   }
  ]
 },
 "generated_utc": "2026-09-09T01:25:26Z"
}