{ "schema_version": "1.1", "instrument": { "figures_read_from": "cite-cut prose (gates.prose_of) since 2026-09-09", "warm_arm_walk": "per page, one warm immediately before that page's own questions, since 2026-09-09", "not_comparable_with": "reports whose schema_version is 1.0" }, "run_utc": "2026-09-12T18:58:31Z", "arm": "warm", "concurrency": 6, "effective_concurrency": 2, "seat": { "url_host_withheld": true, "model": "gemma-4-26b-a4b-fp8", "class": "gemma-class", "seed": "no --seed flag sent" }, "packnorm_definition_recorded_not_governing": "v3", "bank": { "path": "bank.json", "counts": { "on_page": 41, "off_page": 21, "total": 62 } }, "results": { "on_page_total": 41, "on_page_answered": 41, "non_verbatim_figures": 0, "uncited": 0, "off_page_total": 21, "off_page_abstained": 21, "off_page_abstained_by_gate": 1, "off_page_abstained_by_model": 20, "off_page_flagged_but_not_verbatim": [], "on_page_abstained": 1, "off_page_cited": 20, "redaction_hits": 0, "cite_overlap_p50": 0.6185, "cite_overlap_min": 0.0, "cite_overlap_max": 1.0, "cite_overlap_rows": 62, "seat_errors": 0, "p50_seconds": 0.873, "p95_seconds": 1.459, "pages_warmed": 41, "cache_hits": 62, "cache_hit_of": 62 }, "rows": [ { "slug": "three-new-voices-at-the-narrators-chair", "q": "what did the round cost in total, and how were the three new voices judged?", "ok": true, "answer": "The round's metered total came to $0.7652 § {#the-bill}. The three new voices were judged by a panel of seven seats from seven model families, six of which carried the round after one seat was retired for the round § {#seven-seats-six-carried-and-where-their-zeros-sit}. The article does not state how the three new voices were judged specifically, though it notes that the panel read lettered replies blind with no model names on them § {#seven-seats-six-carried-and-where-their-zeros-sit}.", "served": "The round's metered total came to $0.7652 § {#the-bill}. The three new voices were judged by a panel of seven seats from seven model families, six of which carried the round after one seat was retired for the round § {#seven-seats-six-carried-and-where-their-zeros-sit}. The article does not state how the three new voices were judged specifically, though it notes that the panel read lettered replies blind with no model names on them § {#seven-seats-six-carried-and-where-their-zeros-sit}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 2.0628455599653535, "ttft_ms": 674, "tokens_in": 82748, "tokens_out": 125, "cached_tokens": 82464, "cites": [ "the-bill", "seven-seats-six-carried-and-where-their-zeros-sit" ], "gates": { "cites": [ "the-bill", "seven-seats-six-carried-and-where-their-zeros-sit" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "0.7652" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-bill", "overlap": 0.8 }, { "i": 1, "cite": "seven-seats-six-carried-and-where-their-zeros-sit", "overlap": 0.75 }, { "i": 2, "cite": "seven-seats-six-carried-and-where-their-zeros-sit", "overlap": 0.438 } ], "cite_overlap_p50": 0.75, "id": "on-01", "kind": "on-page" }, { "slug": "three-new-voices-at-the-narrators-chair", "q": "what will the next article on this site be about?", "ok": true, "answer": "The article does not say what the next article on the site will be about, though it mentions a previous exhibit called exhibit eleven in § {#come-down-into-the-cove}.", "served": "This page doesn't say. The nearest it gets is § {#come-down-into-the-cove}. qwen3.8:27b across four house benches: one seat filled, one floor missed covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "come-down-into-the-cove", "seconds": 1.3621387319872156, "ttft_ms": 733, "tokens_in": 82742, "tokens_out": 40, "cached_tokens": 82720, "cites": [ "come-down-into-the-cove" ], "gates": { "cites": [ "come-down-into-the-cove" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "come-down-into-the-cove", "overlap": 0.333 } ], "cite_overlap_p50": 0.333, "id": "off-04", "kind": "off-page" }, { "slug": "two-new-frontier-models-at-the-rules-desk", "q": "how many cases did the blind head-to-head use, and did it separate the two models?", "ok": true, "answer": "The blind head-to-head used 36 cases and did not separate the two models § {#head-to-head}. The pooled preference rate was 0.544 for Claude Fable 5.1, and a cluster bootstrap over those cases put that rate at [0.444, 0.638] § {#head-to-head}. The narrator's chair job was moved out of this round before pre-registration was written because it was the most code and the least defensible number on the page § {#two-strangers-at-the-door}.", "served": "The blind head-to-head used 36 cases and did not separate the two models § {#head-to-head}. The pooled preference rate was 0.544 for Claude Fable 5.1, and a cluster bootstrap over those cases put that rate at [0.444, 0.638] § {#head-to-head}. The narrator's chair job was moved out of this round before pre-registration was written because it was the most code and the least defensible number on the page § {#two-strangers-at-the-door}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.3314841479877941, "ttft_ms": 416, "tokens_in": 26867, "tokens_out": 127, "cached_tokens": 26560, "cites": [ "head-to-head", "two-strangers-at-the-door" ], "gates": { "cites": [ "head-to-head", "two-strangers-at-the-door" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "36", "0.544", "5.1", "0.444", "0.638" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "head-to-head", "overlap": 0.875 }, { "i": 1, "cite": "head-to-head", "overlap": 1.0 }, { "i": 2, "cite": "two-strangers-at-the-door", "overlap": 1.0 } ], "cite_overlap_p50": 1.0, "id": "on-02", "kind": "on-page" }, { "slug": "six-worlds-painted-large", "q": "how many worlds were painted, and at what size and step count?", "ok": true, "answer": "Six worlds were painted at 2048×1280 and 40 steps § {#for-the-technically-curious}. The article does not mention any specific figures for what was left out, moved, or not measured § {#for-the-technically-curious}.", "served": "Six worlds were painted at 2048×1280 and 40 steps § {#for-the-technically-curious}. The article does not mention any specific figures for what was left out, moved, or not measured § {#for-the-technically-curious}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.572764418029692, "ttft_ms": 212, "tokens_in": 3099, "tokens_out": 64, "cached_tokens": 2816, "cites": [ "for-the-technically-curious" ], "gates": { "cites": [ "for-the-technically-curious" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "1280", "40" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "for-the-technically-curious", "overlap": 0.857 }, { "i": 1, "cite": "for-the-technically-curious", "overlap": 0.0 } ], "cite_overlap_p50": 0.428, "id": "on-03", "kind": "on-page" }, { "slug": "six-worlds-painted-large", "q": "how many hours a week do the people who ran this work?", "ok": true, "answer": "The article does not say how many hours a week the people who ran this work § {#who-ran-this-and-thanks}.", "served": "This page doesn't say. The nearest it gets is § {#who-ran-this-and-thanks}.", "abstained": true, "abstain_source": "model", "abstain_section": "who-ran-this-and-thanks", "seconds": 0.39703635702608153, "ttft_ms": 218, "tokens_in": 3098, "tokens_out": 29, "cached_tokens": 3072, "cites": [ "who-ran-this-and-thanks" ], "gates": { "cites": [ "who-ran-this-and-thanks" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "who-ran-this-and-thanks", "overlap": 0.143 } ], "cite_overlap_p50": 0.143, "id": "off-10", "kind": "off-page" }, { "slug": "the-beat-lab-asked-and-answered", "q": "how many answers does the page carry, and does any model run to make it?", "ok": true, "answer": "The page carries twenty-seven answers and no model is asked for any of them § {#where-these-answers-come-from}. The answers are derived from a single file in the lab's repository so that an edit to a single word moves both the genie's answer and this page in the same commit § {#where-these-answers-come-from}.", "served": "The page carries twenty-seven answers and no model is asked for any of them § {#where-these-answers-come-from}. The answers are derived from a single file in the lab's repository so that an edit to a single word moves both the genie's answer and this page in the same commit § {#where-these-answers-come-from}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.6413262090063654, "ttft_ms": 234, "tokens_in": 4597, "tokens_out": 77, "cached_tokens": 4320, "cites": [ "where-these-answers-come-from" ], "gates": { "cites": [ "where-these-answers-come-from" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "where-these-answers-come-from", "overlap": 0.833 }, { "i": 1, "cite": "where-these-answers-come-from", "overlap": 0.429 } ], "cite_overlap_p50": 0.631, "id": "on-04", "kind": "on-page" }, { "slug": "ten-minutes-with-living-artists", "q": "how many tracks and releases went into the adapter, and what did the one listener prefer?", "ok": true, "answer": "The adapter was trained on 159 tracks from 41 releases § {#a-genre-not-a-composer}. The one listener preferred the untouched model both times § {#the-ear}. The article does not measure whether the untouched model is any good at this genre § {#what-we-did-not-measure}.", "served": "The adapter was trained on 159 tracks from 41 releases § {#a-genre-not-a-composer}. The one listener preferred the untouched model both times § {#the-ear}. The article does not measure whether the untouched model is any good at this genre § {#what-we-did-not-measure}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7826486519770697, "ttft_ms": 338, "tokens_in": 13160, "tokens_out": 70, "cached_tokens": 12864, "cites": [ "a-genre-not-a-composer", "the-ear", "what-we-did-not-measure" ], "gates": { "cites": [ "a-genre-not-a-composer", "the-ear", "what-we-did-not-measure" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "159", "41" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "a-genre-not-a-composer", "overlap": 0.5 }, { "i": 1, "cite": "the-ear", "overlap": 0.714 }, { "i": 2, "cite": "what-we-did-not-measure", "overlap": 0.833 } ], "cite_overlap_p50": 0.714, "id": "on-05", "kind": "on-page" }, { "slug": "listen-for-yourself", "q": "how many renders are on the page, and what was done to their loudness?", "ok": true, "answer": "There are sixty-three players on the page, each playing a listening master that has had one constant gain applied and nothing else § {#how-to-listen}. The mastered copies are the render with one constant gain applied to reach either −16 LUFS or to respect the −1 dBTP ceiling § {#how-to-listen}. The raw renders themselves were never re-encoded, never trimmed, and never levelled after the fact § {#how-to-listen}.", "served": "There are sixty-three players on the page, each playing a listening master that has had one constant gain applied and nothing else § {#how-to-listen}. The mastered copies are the render with one constant gain applied to reach either −16 LUFS or to respect the −1 dBTP ceiling § {#how-to-listen}. The raw renders themselves were never re-encoded, never trimmed, and never levelled after the fact § {#how-to-listen}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.8868886760319583, "ttft_ms": 369, "tokens_in": 14417, "tokens_out": 97, "cached_tokens": 14112, "cites": [ "how-to-listen" ], "gates": { "cites": [ "how-to-listen" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "16" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "how-to-listen", "overlap": 1.0 }, { "i": 1, "cite": "how-to-listen", "overlap": 0.714 }, { "i": 2, "cite": "how-to-listen", "overlap": 0.889 } ], "cite_overlap_p50": 0.889, "id": "on-06", "kind": "on-page" }, { "slug": "half-an-hour-with-dead-composers", "q": "how long did the Sousa and Bach adapters take to train?", "ok": true, "answer": "The Sousa adapter took 14 minutes 36 seconds to train § {#sousa-a-hundred-and-six-marches}, while the Bach adapter took 8 minutes 23 seconds § {#bach-the-pianist-who-freed-her-own-recordings}. These times were recorded by the trainer's own clock § {#the-bill}. The article does not state what was left out or moved regarding these specific training durations.", "served": "The Sousa adapter took 14 minutes 36 seconds to train § {#sousa-a-hundred-and-six-marches}, while the Bach adapter took 8 minutes 23 seconds § {#bach-the-pianist-who-freed-her-own-recordings}. These times were recorded by the trainer's own clock § {#the-bill}. The article does not state what was left out or moved regarding these specific training durations.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7848168730270118, "ttft_ms": 248, "tokens_in": 5657, "tokens_out": 98, "cached_tokens": 5376, "cites": [ "sousa-a-hundred-and-six-marches", "bach-the-pianist-who-freed-her-own-recordings", "the-bill" ], "gates": { "cites": [ "sousa-a-hundred-and-six-marches", "bach-the-pianist-who-freed-her-own-recordings", "the-bill" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "14", "36", "8", "23" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "sousa-a-hundred-and-six-marches", "overlap": 0.545 }, { "i": 0, "cite": "bach-the-pianist-who-freed-her-own-recordings", "overlap": 0.455 }, { "i": 1, "cite": "the-bill", "overlap": 0.6 } ], "cite_overlap_p50": 0.545, "id": "on-07", "kind": "on-page" }, { "slug": "chopin-in-five-minutes", "q": "how many recordings were used, and how long did the training take by the trainer's clock?", "ok": true, "answer": "Nineteen recordings were used for training, which took 4 minutes 49 seconds by the trainer's own clock § {#five-minutes-of-training-honestly-accounted}. The article notes that it did not measure whether a smaller GPU could have run this training or what happens past ten passes § {#what-we-did-not-measure}.", "served": "Nineteen recordings were used for training, which took 4 minutes 49 seconds by the trainer's own clock § {#five-minutes-of-training-honestly-accounted}. The article notes that it did not measure whether a smaller GPU could have run this training or what happens past ten passes § {#what-we-did-not-measure}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.6214964659884572, "ttft_ms": 227, "tokens_in": 3797, "tokens_out": 76, "cached_tokens": 3488, "cites": [ "five-minutes-of-training-honestly-accounted", "what-we-did-not-measure" ], "gates": { "cites": [ "five-minutes-of-training-honestly-accounted", "what-we-did-not-measure" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "4", "49" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "five-minutes-of-training-honestly-accounted", "overlap": 0.818 }, { "i": 1, "cite": "what-we-did-not-measure", "overlap": 0.727 } ], "cite_overlap_p50": 0.772, "id": "on-08", "kind": "on-page" }, { "slug": "the-same-sixteen", "q": "what did glm-5.3-flash score on the tool bench, and what did the local models score?", "ok": true, "answer": "The reference arm scored 16/19 — 84.2%, Wilson 95% interval [62.4%, 94.5%] on the tool bench § {#sixteen-of-nineteen}. Two local models, gemma4:12b and qwen3.6:27b, also landed on the exact count of 16/19 § {#two-local-models-sixteen-days-earlier}. The article notes that the judge exam, assistant exam, decode speed, and filing cabinet could not honestly be read on the cloud path § {#the-ledger-of-what-did-not-read}.", "served": "The reference arm scored 16/19 — 84.2%, Wilson 95% interval [62.4%, 94.5%] on the tool bench § {#sixteen-of-nineteen}. Two local models, gemma4:12b and qwen3.6:27b, also landed on the exact count of 16/19 § {#two-local-models-sixteen-days-earlier}. The article notes that the judge exam, assistant exam, decode speed, and filing cabinet could not honestly be read on the cloud path § {#the-ledger-of-what-did-not-read}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0763042379985563, "ttft_ms": 267, "tokens_in": 6507, "tokens_out": 143, "cached_tokens": 6208, "cites": [ "sixteen-of-nineteen", "two-local-models-sixteen-days-earlier", "the-ledger-of-what-did-not-read" ], "gates": { "cites": [ "sixteen-of-nineteen", "two-local-models-sixteen-days-earlier", "the-ledger-of-what-did-not-read" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "16/19", "84.2", "95", "62.4", "94.5", "4:12", "3.6", "27", "16/19" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "sixteen-of-nineteen", "overlap": 0.917 }, { "i": 1, "cite": "two-local-models-sixteen-days-earlier", "overlap": 1.0 }, { "i": 2, "cite": "the-ledger-of-what-did-not-read", "overlap": 0.833 } ], "cite_overlap_p50": 0.917, "id": "on-09", "kind": "on-page" }, { "slug": "the-same-sixteen", "q": "which of these models is the best one to use for my own project?", "ok": true, "answer": "The article does not answer which model is the best to use for a project § {#a-different-kind-of-row}.", "served": "This page doesn't say. The nearest it gets is § {#a-different-kind-of-row}. Three New Voices at the Narrator's Chair covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "a-different-kind-of-row", "seconds": 0.4607623359770514, "ttft_ms": 273, "tokens_in": 6499, "tokens_out": 28, "cached_tokens": 6464, "cites": [ "a-different-kind-of-row" ], "gates": { "cites": [ "a-different-kind-of-row" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "a-different-kind-of-row", "overlap": 0.4 } ], "cite_overlap_p50": 0.4, "id": "off-20", "kind": "off-page" }, { "slug": "two-hours-on-battery", "q": "what speed did the mini pc hold on battery, and for how long?", "ok": true, "answer": "The mini PC held a speed of ~12 tokens/second on battery for about two hours § {#two-hours-at-12-tok-s-the-test-that-named-the-page}. The article notes that it did not measure battery-vs-plugged behavior for the GPU-server half, sustained thermals past the measured window, or non-Q4 quantizations on CPU § {#what-we-did-not-measure}.", "served": "The mini PC held a speed of ~12 tokens/second on battery for about two hours § {#two-hours-at-12-tok-s-the-test-that-named-the-page}. The article notes that it did not measure battery-vs-plugged behavior for the GPU-server half, sustained thermals past the measured window, or non-Q4 quantizations on CPU § {#what-we-did-not-measure}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0246263309963979, "ttft_ms": 341, "tokens_in": 11048, "tokens_out": 98, "cached_tokens": 10752, "cites": [ "two-hours-at-12-tok-s-the-test-that-named-the-page", "what-we-did-not-measure" ], "gates": { "cites": [ "two-hours-at-12-tok-s-the-test-that-named-the-page", "what-we-did-not-measure" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "12", "4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "two-hours-at-12-tok-s-the-test-that-named-the-page", "overlap": 1.0 }, { "i": 1, "cite": "what-we-did-not-measure", "overlap": 0.786 } ], "cite_overlap_p50": 0.893, "id": "on-10", "kind": "on-page" }, { "slug": "two-hours-on-battery", "q": "what did this cost in electricity at my own utility's rate?", "ok": true, "answer": "The article does not provide information regarding the cost of electricity at a user's own utility rate § {#two-hours-at-12-tok-s-the-test-that-named-the-page}. It only notes the power draw in watts during various phases of the test § {#two-hours-at-12-tok-s-the-test-that-named-the-page}.", "served": "This page doesn't say. The nearest it gets is § {#two-hours-at-12-tok-s-the-test-that-named-the-page}.", "abstained": true, "abstain_source": "model", "abstain_section": "two-hours-at-12-tok-s-the-test-that-named-the-page", "seconds": 0.9448192290146835, "ttft_ms": 354, "tokens_in": 11047, "tokens_out": 88, "cached_tokens": 11008, "cites": [ "two-hours-at-12-tok-s-the-test-that-named-the-page" ], "gates": { "cites": [ "two-hours-at-12-tok-s-the-test-that-named-the-page" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "two-hours-at-12-tok-s-the-test-that-named-the-page", "overlap": 0.111 }, { "i": 1, "cite": "two-hours-at-12-tok-s-the-test-that-named-the-page", "overlap": 0.333 } ], "cite_overlap_p50": 0.222, "id": "off-05", "kind": "off-page" }, { "slug": "the-instrument-travels", "q": "when the frozen exam moved to the smaller card, where did the ranked model land?", "ok": true, "answer": "The ranked model, qwen3.6:27b, landed at the floor by clearing both judge floors with a kill-recall of 11/12 and a preservation of 8/9 § {#the-instrument-travels}. This result was achieved on a different build and different silicon than the previous run § {#the-instrument-travels}. The article notes that the ninth preservation item for this model was unscorable rather than lost, and that the battery stopped itself mid-leg when free space fell below the 60 GiB safety floor § {#what-we-ran-and-the-rule-behind-every-number}.", "served": "The ranked model, qwen3.6:27b, landed at the floor by clearing both judge floors with a kill-recall of 11/12 and a preservation of 8/9 § {#the-instrument-travels}. This result was achieved on a different build and different silicon than the previous run § {#the-instrument-travels}. The article notes that the ninth preservation item for this model was unscorable rather than lost, and that the battery stopped itself mid-leg when free space fell below the 60 GiB safety floor § {#what-we-ran-and-the-rule-behind-every-number}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0486706860247068, "ttft_ms": 264, "tokens_in": 8731, "tokens_out": 137, "cached_tokens": 8448, "cites": [ "what-we-ran-and-the-rule-behind-every-number" ], "gates": { "cites": [ "what-we-ran-and-the-rule-behind-every-number" ], "cites_dropped": [ "the-instrument-travels" ], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [ "cite: 1 cite(s) not in this page's anchors: the-instrument-travels" ], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [ "the-instrument-travels" ] }, "figures": [ "3.6", "27", "11/12", "8/9", "60" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 2, "cite": "what-we-ran-and-the-rule-behind-every-number", "overlap": 0.7 } ], "cite_overlap_p50": 0.7, "id": "on-11", "kind": "on-page" }, { "slug": "nine-worlds-one-dog", "q": "what stays the same across the nine paintings?", "ok": true, "answer": "The founder's dog remains the same, featuring the same curl and tucked paws, while napping under the same tree in every world § {#top}. The dog is a large black Portuguese Water Dog with one bright white chest patch §. The article does not mention anything that was left out, moved, or not measured.", "served": "The founder's dog remains the same, featuring the same curl and tucked paws, while napping under the same tree in every world § {#top}. The dog is a large black Portuguese Water Dog with one bright white chest patch §. The article does not mention anything that was left out, moved, or not measured.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.5362216340145096, "ttft_ms": 184, "tokens_in": 2012, "tokens_out": 66, "cached_tokens": 1728, "cites": [ "top" ], "gates": { "cites": [ "top" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "top", "overlap": 0.5 } ], "cite_overlap_p50": 0.5, "id": "on-12", "kind": "on-page" }, { "slug": "nine-worlds-one-dog", "q": "how many other readers have asked this same question?", "ok": true, "answer": "The article does not answer how many other readers have asked this same question § {#for-the-technically-curious}.", "served": "This page doesn't say. The nearest it gets is § {#for-the-technically-curious}.", "abstained": true, "abstain_source": "model", "abstain_section": "for-the-technically-curious", "seconds": 0.34557090199086815, "ttft_ms": 186, "tokens_in": 2013, "tokens_out": 27, "cached_tokens": 1984, "cites": [ "for-the-technically-curious" ], "gates": { "cites": [ "for-the-technically-curious" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "for-the-technically-curious", "overlap": 0.286 } ], "cite_overlap_p50": 0.286, "id": "off-17", "kind": "off-page" }, { "slug": "what-150-watts-buys", "q": "what did the 450-watt cap cost the dense model, and what did it buy back?", "ok": true, "answer": "The 600 to 450 W cap cost the dense model 2.0% of throughput and bought back 138 W, 11 °C, and a 28% jump in efficiency § {#the-answer-in-one-paragraph}. At concurrency 16, the throughput moved from 94.6 to 92.7 tok/s per stream § {#the-answer-in-one-paragraph}. The article notes that it did not measure the cost for long-context or prefill-heavy workloads § {#what-this-page-does-not-say}.", "served": "The 600 to 450 W cap cost the dense model 2.0% of throughput and bought back 138 W, 11 °C, and a 28% jump in efficiency § {#the-answer-in-one-paragraph}. At concurrency 16, the throughput moved from 94.6 to 92.7 tok/s per stream § {#the-answer-in-one-paragraph}. The article notes that it did not measure the cost for long-context or prefill-heavy workloads § {#what-this-page-does-not-say}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0534670879715122, "ttft_ms": 273, "tokens_in": 7223, "tokens_out": 132, "cached_tokens": 6912, "cites": [ "the-answer-in-one-paragraph", "what-this-page-does-not-say" ], "gates": { "cites": [ "the-answer-in-one-paragraph", "what-this-page-does-not-say" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "600", "450", "2.0", "138", "11", "28", "16", "94.6", "92.7" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-answer-in-one-paragraph", "overlap": 0.8 }, { "i": 1, "cite": "the-answer-in-one-paragraph", "overlap": 1.0 }, { "i": 2, "cite": "what-this-page-does-not-say", "overlap": 0.5 } ], "cite_overlap_p50": 0.8, "id": "on-13", "kind": "on-page" }, { "slug": "what-150-watts-buys", "q": "what does a large cloud provider charge for the same work?", "ok": true, "answer": "The article does not answer what a large cloud provider charges for the same work § {#the-dial-and-who-notices}.", "served": "This page doesn't say. The nearest it gets is § {#the-dial-and-who-notices}.", "abstained": true, "abstain_source": "model", "abstain_section": "the-dial-and-who-notices", "seconds": 0.47429355897475034, "ttft_ms": 282, "tokens_in": 7213, "tokens_out": 29, "cached_tokens": 7168, "cites": [ "the-dial-and-who-notices" ], "gates": { "cites": [ "the-dial-and-who-notices" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-dial-and-who-notices", "overlap": 0.286 } ], "cite_overlap_p50": 0.286, "id": "off-01", "kind": "off-page" }, { "slug": "how-the-beat-lab-works", "q": "where is the sound made, and how long does a typical wish take?", "ok": true, "answer": "Every sound in the Beat Lab is synthesised by the user's own browser § {#the-sound-is-made-on-your-device}. A typical wish takes about 4 s § {#what-happens-when-you-ask-the-genie}. The article does not mention what was left out, moved, or not measured in relation to these specific details.", "served": "Every sound in the Beat Lab is synthesised by the user's own browser § {#the-sound-is-made-on-your-device}. A typical wish takes about 4 s § {#what-happens-when-you-ask-the-genie}. The article does not mention what was left out, moved, or not measured in relation to these specific details.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.219647502002772, "ttft_ms": 489, "tokens_in": 31423, "tokens_out": 81, "cached_tokens": 31136, "cites": [ "the-sound-is-made-on-your-device", "what-happens-when-you-ask-the-genie" ], "gates": { "cites": [ "the-sound-is-made-on-your-device", "what-happens-when-you-ask-the-genie" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-sound-is-made-on-your-device", "overlap": 0.875 }, { "i": 1, "cite": "what-happens-when-you-ask-the-genie", "overlap": 1.0 } ], "cite_overlap_p50": 0.938, "id": "on-14", "kind": "on-page" }, { "slug": "how-the-beat-lab-works", "q": "what phone number reaches support?", "ok": true, "answer": "The article does not provide a phone number to reach support § {#what-it-is}.", "served": "This page doesn't say. The nearest it gets is § {#what-it-is}.", "abstained": true, "abstain_source": "model", "abstain_section": "what-it-is", "seconds": 0.842417796025984, "ttft_ms": 513, "tokens_in": 31414, "tokens_out": 20, "cached_tokens": 31392, "cites": [ "what-it-is" ], "gates": { "cites": [ "what-it-is" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "what-it-is", "overlap": 0.0 } ], "cite_overlap_p50": 0.0, "id": "off-18", "kind": "off-page" }, { "slug": "the-typist-and-the-developer", "q": "how long did the developer take to develop one picture?", "ok": true, "answer": "The developer developed a picture in 1.54 seconds § {#the-developer}. When the typist was streaming tokens at the same time, the developer took 1.82 seconds per sketch § {#one-card-both-jobs-at-once}. The article notes that it did not instrument how much traffic was present during the run § {#what-we-did-not-measure-said-out-loud}.", "served": "The developer developed a picture in 1.54 seconds § {#the-developer}. When the typist was streaming tokens at the same time, the developer took 1.82 seconds per sketch § {#one-card-both-jobs-at-once}. The article notes that it did not instrument how much traffic was present during the run § {#what-we-did-not-measure-said-out-loud}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.8002599229803309, "ttft_ms": 258, "tokens_in": 5490, "tokens_out": 90, "cached_tokens": 5216, "cites": [ "the-developer", "one-card-both-jobs-at-once", "what-we-did-not-measure-said-out-loud" ], "gates": { "cites": [ "the-developer", "one-card-both-jobs-at-once", "what-we-did-not-measure-said-out-loud" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "1.54", "1.82" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-developer", "overlap": 1.0 }, { "i": 1, "cite": "one-card-both-jobs-at-once", "overlap": 0.909 }, { "i": 2, "cite": "what-we-did-not-measure-said-out-loud", "overlap": 0.714 } ], "cite_overlap_p50": 0.909, "id": "on-15", "kind": "on-page" }, { "slug": "the-typist-and-the-developer", "q": "what is this workshop's annual revenue?", "ok": true, "answer": "The article does not say what this workshop's annual revenue is, though the section § {#one-card-both-jobs-at-once} discusses the estate context.", "served": "This page doesn't say. The nearest it gets is § {#one-card-both-jobs-at-once}.", "abstained": true, "abstain_source": "model", "abstain_section": "one-card-both-jobs-at-once", "seconds": 0.5219879529904574, "ttft_ms": 265, "tokens_in": 5488, "tokens_out": 37, "cached_tokens": 5440, "cites": [ "one-card-both-jobs-at-once" ], "gates": { "cites": [ "one-card-both-jobs-at-once" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "one-card-both-jobs-at-once", "overlap": 0.333 } ], "cite_overlap_p50": 0.333, "id": "off-12", "kind": "off-page" }, { "slug": "a-rig-your-friend-already-owns", "q": "how much faster was the workstation card than the 3090 on the game's own art?", "ok": true, "answer": "The article does not provide a single speed comparison for the workstation card against the 3090 for the game's own art, instead reporting median seconds per render for various models § {#the-table-that-answers-the-question}. For the game's sketch artist, FLUX.2 klein-4B, the 96G card was 3.4× faster than the 3090 § {#the-table-that-answers-the-question}. Other models showed different ratios, such as z-image-turbo being 3.9× slower on the 3090 and krea2-turbo being 4.9× slower § {#the-table-that-answers-the-question}. The text notes that 8–12 GB cards were not tested § {#what-this-means-for-the-game}.", "served": "This page doesn't say. The nearest it gets is § {#the-table-that-answers-the-question}. Fifty-Seven Milliseconds: What a Diffusion Step Actually Buys covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "the-table-that-answers-the-question", "seconds": 1.2569750039838254, "ttft_ms": 252, "tokens_in": 4731, "tokens_out": 185, "cached_tokens": 4448, "cites": [ "the-table-that-answers-the-question", "what-this-means-for-the-game" ], "gates": { "cites": [ "the-table-that-answers-the-question", "what-this-means-for-the-game" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "3090", "2", "4", "96", "3.4", "3090", "3.9", "3090", "2", "4.9", "8-12" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-table-that-answers-the-question", "overlap": 0.421 }, { "i": 1, "cite": "the-table-that-answers-the-question", "overlap": 0.8 }, { "i": 2, "cite": "the-table-that-answers-the-question", "overlap": 0.538 }, { "i": 3, "cite": "what-this-means-for-the-game", "overlap": 0.333 } ], "cite_overlap_p50": 0.48, "id": "on-16", "kind": "on-page" }, { "slug": "a-rig-your-friend-already-owns", "q": "what does one of these cards cost in my country today?", "ok": true, "answer": "The article does not answer what the cost of these cards is in the reader's country today § {#the-moment-this-page-is-about}.", "served": "This page doesn't say. The nearest it gets is § {#the-moment-this-page-is-about}. Three New Voices at the Narrator's Chair covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "the-moment-this-page-is-about", "seconds": 0.4718162270146422, "ttft_ms": 257, "tokens_in": 4721, "tokens_out": 34, "cached_tokens": 4672, "cites": [ "the-moment-this-page-is-about" ], "gates": { "cites": [ "the-moment-this-page-is-about" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-moment-this-page-is-about", "overlap": 0.167 } ], "cite_overlap_p50": 0.167, "id": "off-07", "kind": "off-page" }, { "slug": "how-a-vision-model-sees", "q": "how many tokens does an 896 px picture become, and how long does its prefill take?", "ok": true, "answer": "An 896 px picture becomes 199 tokens and its prefill takes 184 ms § {#finding-one-a-picture-is-199-tokens-at-896-px-and-the}. This measurement was performed on the seat the cove ships, which is the minicpm-v4.5 model § {#finding-one-a-picture-is-199-tokens-at-896-px-and-the}.", "served": "An 896 px picture becomes 199 tokens and its prefill takes 184 ms § {#finding-one-a-picture-is-199-tokens-at-896-px-and-the}. This measurement was performed on the seat the cove ships, which is the minicpm-v4.5 model § {#finding-one-a-picture-is-199-tokens-at-896-px-and-the}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.9596257790108211, "ttft_ms": 315, "tokens_in": 10265, "tokens_out": 107, "cached_tokens": 9984, "cites": [ "finding-one-a-picture-is-199-tokens-at-896-px-and-the" ], "gates": { "cites": [ "finding-one-a-picture-is-199-tokens-at-896-px-and-the" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "896", "199", "184", "4.5" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "finding-one-a-picture-is-199-tokens-at-896-px-and-the", "overlap": 0.8 }, { "i": 1, "cite": "finding-one-a-picture-is-199-tokens-at-896-px-and-the", "overlap": 0.714 } ], "cite_overlap_p50": 0.757, "id": "on-17", "kind": "on-page" }, { "slug": "fifty-seven-milliseconds", "q": "which painter is the fastest cell, and at what time?", "ok": true, "answer": "The distilled painter is the sharpest and fastest cell in the whole second run, recorded at 0.62 s at 4 steps § {#finding-four-the-argument-with-our-own-settings-and-what-it}. The article notes that the published pages serve compressed derivatives rather than the lossless masters used for the measurements § {#how-to-check-our-work-and-see-it-live}.", "served": "The distilled painter is the sharpest and fastest cell in the whole second run, recorded at 0.62 s at 4 steps § {#finding-four-the-argument-with-our-own-settings-and-what-it}. The article notes that the published pages serve compressed derivatives rather than the lossless masters used for the measurements § {#how-to-check-our-work-and-see-it-live}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7981019000289962, "ttft_ms": 260, "tokens_in": 5388, "tokens_out": 92, "cached_tokens": 5088, "cites": [ "finding-four-the-argument-with-our-own-settings-and-what-it", "how-to-check-our-work-and-see-it-live" ], "gates": { "cites": [ "finding-four-the-argument-with-our-own-settings-and-what-it", "how-to-check-our-work-and-see-it-live" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "0.62", "4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "finding-four-the-argument-with-our-own-settings-and-what-it", "overlap": 0.909 }, { "i": 1, "cite": "how-to-check-our-work-and-see-it-live", "overlap": 0.636 } ], "cite_overlap_p50": 0.772, "id": "on-18", "kind": "on-page" }, { "slug": "fifty-seven-milliseconds", "q": "what was the weather like on the day this was measured?", "ok": true, "answer": "The article does not say what the weather was like, though it mentions the first run was measured on 2026-07-10 § {#finding-one-a-step-is-a-fixed-price-good}.", "served": "This page doesn't say. The nearest it gets is § {#finding-one-a-step-is-a-fixed-price-good}. What 150 Watts Buys covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "finding-one-a-step-is-a-fixed-price-good", "seconds": 0.5909565020119771, "ttft_ms": 265, "tokens_in": 5388, "tokens_out": 52, "cached_tokens": 5344, "cites": [ "finding-one-a-step-is-a-fixed-price-good" ], "gates": { "cites": [ "finding-one-a-step-is-a-fixed-price-good" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "2026-07-10" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "finding-one-a-step-is-a-fixed-price-good", "overlap": 0.444 } ], "cite_overlap_p50": 0.444, "id": "off-09", "kind": "off-page" }, { "slug": "the-dog-the-dice-and-the-painter", "q": "how big is the file the whole world is written in?", "ok": true, "answer": "The world is written in a single YAML file that is 16,705 lines § {#the-world-is-a-file}. This file contains six regions, forty NPCs, fifty-nine quests, and three kinds of adoptable pet § {#the-world-is-a-file}. The article does not mention the file's size in bytes, though it notes that two copies are byte-identical § {#the-world-is-a-file}.", "served": "The world is written in a single YAML file that is 16,705 lines § {#the-world-is-a-file}. This file contains six regions, forty NPCs, fifty-nine quests, and three kinds of adoptable pet § {#the-world-is-a-file}. The article does not mention the file's size in bytes, though it notes that two copies are byte-identical § {#the-world-is-a-file}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7938549209502526, "ttft_ms": 251, "tokens_in": 6070, "tokens_out": 100, "cached_tokens": 5792, "cites": [ "the-world-is-a-file" ], "gates": { "cites": [ "the-world-is-a-file" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "16,705" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-world-is-a-file", "overlap": 0.875 }, { "i": 1, "cite": "the-world-is-a-file", "overlap": 0.917 }, { "i": 2, "cite": "the-world-is-a-file", "overlap": 0.333 } ], "cite_overlap_p50": 0.875, "id": "on-19", "kind": "on-page" }, { "slug": "where-new-knowledge-comes-from", "q": "what are the two engines the page names for where new knowledge comes from?", "ok": true, "answer": "The two engines named are present in the data, unmined, and implied by the rules, unvisited § {#two-kinds-of-there-all-along}. The first engine involves reading patterns already existing in archives, such as co-evolution fingerprints or the structure-activity patterns in a library of 6,111 shelved compounds § {#the-first-engine-the-archive-nobody-had-read}. The second engine involves a rule that captures a generating rule to emit entries that no row of data contains, such as a planet's position in 1631 § {#the-second-engine-the-rooms-the-rules-imply}. Structural biologists noted that these engines left out how proteins pair up, how they move, and what one mutation does § {#the-first-engine-the-archive-nobody-had-read}.", "served": "The two engines named are present in the data, unmined, and implied by the rules, unvisited § {#two-kinds-of-there-all-along}. The first engine involves reading patterns already existing in archives, such as co-evolution fingerprints or the structure-activity patterns in a library of 6,111 shelved compounds § {#the-first-engine-the-archive-nobody-had-read}. The second engine involves a rule that captures a generating rule to emit entries that no row of data contains, such as a planet's position in 1631 § {#the-second-engine-the-rooms-the-rules-imply}. Structural biologists noted that these engines left out how proteins pair up, how they move, and what one mutation does § {#the-first-engine-the-archive-nobody-had-read}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.2301901149912737, "ttft_ms": 250, "tokens_in": 5426, "tokens_out": 183, "cached_tokens": 5152, "cites": [ "two-kinds-of-there-all-along", "the-first-engine-the-archive-nobody-had-read", "the-second-engine-the-rooms-the-rules-imply" ], "gates": { "cites": [ "two-kinds-of-there-all-along", "the-first-engine-the-archive-nobody-had-read", "the-second-engine-the-rooms-the-rules-imply" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "6,111", "1631" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "two-kinds-of-there-all-along", "overlap": 0.889 }, { "i": 1, "cite": "the-first-engine-the-archive-nobody-had-read", "overlap": 0.438 }, { "i": 2, "cite": "the-second-engine-the-rooms-the-rules-imply", "overlap": 0.6 }, { "i": 3, "cite": "the-first-engine-the-archive-nobody-had-read", "overlap": 0.231 } ], "cite_overlap_p50": 0.519, "id": "on-20", "kind": "on-page" }, { "slug": "compression-is-the-objective", "q": "what does the page say a language model is, and what prize does it point to?", "ok": true, "answer": "A language model is a compression of its corpus § {#the-punchline}. The article points to the Hutter Prize, which charges for everything including self-extracting entries and every decompressor byte counted § {#the-punchline}. The text notes that the scoreboard for the 2023 paper \"Language Modeling Is Compression\" never charges for the model's own weights § {#the-punchline}.", "served": "A language model is a compression of its corpus § {#the-punchline}. The article points to the Hutter Prize, which charges for everything including self-extracting entries and every decompressor byte counted § {#the-punchline}. The text notes that the scoreboard for the 2023 paper \"Language Modeling Is Compression\" never charges for the model's own weights § {#the-punchline}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7715070989797823, "ttft_ms": 283, "tokens_in": 5402, "tokens_out": 86, "cached_tokens": 5344, "cites": [ "the-punchline" ], "gates": { "cites": [ "the-punchline" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-punchline", "overlap": 1.0 }, { "i": 1, "cite": "the-punchline", "overlap": 0.333 }, { "i": 2, "cite": "the-punchline", "overlap": 0.538 } ], "cite_overlap_p50": 0.538, "id": "on-21", "kind": "on-page" }, { "slug": "compression-is-the-objective", "q": "how would these numbers change on windows instead?", "ok": true, "answer": "The article does not say how these numbers would change on windows § {#what-we-can-show-you}.", "served": "This page doesn't say. The nearest it gets is § {#what-we-can-show-you}.", "abstained": true, "abstain_source": "model", "abstain_section": "what-we-can-show-you", "seconds": 0.43877313303528354, "ttft_ms": 266, "tokens_in": 5393, "tokens_out": 25, "cached_tokens": 5120, "cites": [ "what-we-can-show-you" ], "gates": { "cites": [ "what-we-can-show-you" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "what-we-can-show-you", "overlap": 0.25 } ], "cite_overlap_p50": 0.25, "id": "off-13", "kind": "off-page" }, { "slug": "the-free-speed-wasnt-free", "q": "what was the production median before the change, and what caused the speed-up?", "ok": true, "answer": "The production median was 138.6 tok/s over 1,982 live rulings § {#the-day-the-box-got-faster}. The speed-up was caused by an ollama commit that changed the default repeat_penalty from 1.1 to 1.0, which removed a penalties stage that had been running on every token § {#the-two-line-commit}. The article notes that the quality half of this change remains unmeasured § {#the-part-that-isnt-about-speed}.", "served": "The production median was 138.6 tok/s over 1,982 live rulings § {#the-day-the-box-got-faster}. The speed-up was caused by an ollama commit that changed the default repeat_penalty from 1.1 to 1.0, which removed a penalties stage that had been running on every token § {#the-two-line-commit}. The article notes that the quality half of this change remains unmeasured § {#the-part-that-isnt-about-speed}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.910048864025157, "ttft_ms": 260, "tokens_in": 7483, "tokens_out": 116, "cached_tokens": 7200, "cites": [ "the-day-the-box-got-faster", "the-two-line-commit", "the-part-that-isnt-about-speed" ], "gates": { "cites": [ "the-day-the-box-got-faster", "the-two-line-commit", "the-part-that-isnt-about-speed" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "138.6", "1,982", "1.1", "1.0" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-day-the-box-got-faster", "overlap": 1.0 }, { "i": 1, "cite": "the-two-line-commit", "overlap": 0.625 }, { "i": 2, "cite": "the-part-that-isnt-about-speed", "overlap": 0.667 } ], "cite_overlap_p50": 0.667, "id": "on-22", "kind": "on-page" }, { "slug": "the-compressed-photograph", "q": "how much slower was the uncompressed original than the compressed build?", "ok": true, "answer": "The uncompressed original was 3.4× the speed of the trained-4-bit build § {#the-same-question-down-the-whole-ladder}. In the 08-19 pair, the q8 build was 28% slower in tokens per second than the q4 build § {#the-same-brain-at-two-compressions}. The article does not explicitly state the exact speed difference between the uncompressed original and the q4 build, though it provides individual median speeds of 58.1 tok/s and 129.4 tok/s § {#the-same-question-down-the-whole-ladder}. The text notes that it does not measure perplexity § {#what-this-page-does-not-say}.", "served": "The uncompressed original was 3.4× the speed of the trained-4-bit build § {#the-same-question-down-the-whole-ladder}. In the 08-19 pair, the q8 build was 28% slower in tokens per second than the q4 build § {#the-same-brain-at-two-compressions}. The article does not explicitly state the exact speed difference between the uncompressed original and the q4 build, though it provides individual median speeds of 58.1 tok/s and 129.4 tok/s § {#the-same-question-down-the-whole-ladder}. The text notes that it does not measure perplexity § {#what-this-page-does-not-say}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.412406507995911, "ttft_ms": 265, "tokens_in": 8591, "tokens_out": 167, "cached_tokens": 8288, "cites": [ "the-same-question-down-the-whole-ladder", "the-same-brain-at-two-compressions", "what-this-page-does-not-say" ], "gates": { "cites": [ "the-same-question-down-the-whole-ladder", "the-same-brain-at-two-compressions", "what-this-page-does-not-say" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "3.4", "4", "08-19", "8", "28", "4", "4", "58.1", "129.4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-same-question-down-the-whole-ladder", "overlap": 1.0 }, { "i": 1, "cite": "the-same-brain-at-two-compressions", "overlap": 0.8 }, { "i": 2, "cite": "the-same-question-down-the-whole-ladder", "overlap": 0.5 }, { "i": 3, "cite": "what-this-page-does-not-say", "overlap": 0.75 } ], "cite_overlap_p50": 0.775, "id": "on-23", "kind": "on-page" }, { "slug": "three-at-the-table", "q": "what is the true active share for the live seat, and what does the page call the two numbers in a model name?", "ok": true, "answer": "The true active share for the live seat is 14.4% of the model's bytes § {#the-payroll-and-the-meeting}. The two numbers in a model name are the payroll and the meeting § {#the-tag-with-two-numbers}. The article notes that it does not measure the latency trade-off of streaming experts from system RAM § {#the-part-nobody-warns-you-about-the-whole-payroll-needs-an}.", "served": "The true active share for the live seat is 14.4% of the model's bytes § {#the-payroll-and-the-meeting}. The two numbers in a model name are the payroll and the meeting § {#the-tag-with-two-numbers}. The article notes that it does not measure the latency trade-off of streaming experts from system RAM § {#the-part-nobody-warns-you-about-the-whole-payroll-needs-an}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.8949841409921646, "ttft_ms": 280, "tokens_in": 7639, "tokens_out": 103, "cached_tokens": 7328, "cites": [ "the-payroll-and-the-meeting", "the-tag-with-two-numbers", "the-part-nobody-warns-you-about-the-whole-payroll-needs-an" ], "gates": { "cites": [ "the-payroll-and-the-meeting", "the-tag-with-two-numbers", "the-part-nobody-warns-you-about-the-whole-payroll-needs-an" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "14.4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-payroll-and-the-meeting", "overlap": 1.0 }, { "i": 1, "cite": "the-tag-with-two-numbers", "overlap": 0.833 }, { "i": 2, "cite": "the-part-nobody-warns-you-about-the-whole-payroll-needs-an", "overlap": 0.5 } ], "cite_overlap_p50": 0.833, "id": "on-24", "kind": "on-page" }, { "slug": "three-at-the-table", "q": "what would this measurement look like at twice the batch size?", "ok": true, "answer": "The article does not answer what the measurement would look like at twice the batch size § {#measured-on-our-own-machines}.", "served": "This page doesn't say. The nearest it gets is § {#measured-on-our-own-machines}.", "abstained": true, "abstain_source": "model", "abstain_section": "measured-on-our-own-machines", "seconds": 0.48298436898039654, "ttft_ms": 283, "tokens_in": 7626, "tokens_out": 29, "cached_tokens": 7584, "cites": [ "measured-on-our-own-machines" ], "gates": { "cites": [ "measured-on-our-own-machines" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "measured-on-our-own-machines", "overlap": 0.286 } ], "cite_overlap_p50": 0.286, "id": "off-19", "kind": "off-page" }, { "slug": "reading-is-fast", "q": "what was the median reading speed, and why is reading faster than writing?", "ok": true, "answer": "The median reading speed was 5,415 tokens/s § {#a-measurement-from-our-own-card}. Reading is faster because it is parallel, as the model can process all tokens in a prompt at the same time since they already exist § {#reading-is-a-group-photo}. Writing is slower because it is serial, as each new token depends on the tokens that came before it and must be chosen one at a time § {#writing-is-one-at-a-time}. The article notes that it does not say the ratio is a constant § {#what-this-page-does-not-say}.", "served": "The median reading speed was 5,415 tokens/s § {#a-measurement-from-our-own-card}. Reading is faster because it is parallel, as the model can process all tokens in a prompt at the same time since they already exist § {#reading-is-a-group-photo}. Writing is slower because it is serial, as each new token depends on the tokens that came before it and must be chosen one at a time § {#writing-is-one-at-a-time}. The article notes that it does not say the ratio is a constant § {#what-this-page-does-not-say}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.008244989963714, "ttft_ms": 260, "tokens_in": 5511, "tokens_out": 136, "cached_tokens": 5216, "cites": [ "a-measurement-from-our-own-card", "reading-is-a-group-photo", "writing-is-one-at-a-time", "what-this-page-does-not-say" ], "gates": { "cites": [ "a-measurement-from-our-own-card", "reading-is-a-group-photo", "writing-is-one-at-a-time", "what-this-page-does-not-say" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "5,415" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "a-measurement-from-our-own-card", "overlap": 1.0 }, { "i": 1, "cite": "reading-is-a-group-photo", "overlap": 0.786 }, { "i": 2, "cite": "writing-is-one-at-a-time", "overlap": 0.667 }, { "i": 3, "cite": "what-this-page-does-not-say", "overlap": 0.75 } ], "cite_overlap_p50": 0.768, "id": "on-25", "kind": "on-page" }, { "slug": "reading-is-fast", "q": "who won the football world cup in 2022?", "ok": true, "answer": "The article does not answer the question § {#the-mystery-stated-honestly}.", "served": "This page doesn't say. The nearest it gets is § {#the-mystery-stated-honestly}.", "abstained": true, "abstain_source": "model", "abstain_section": "the-mystery-stated-honestly", "seconds": 0.39271158998599276, "ttft_ms": 265, "tokens_in": 5509, "tokens_out": 19, "cached_tokens": 5472, "cites": [ "the-mystery-stated-honestly" ], "gates": { "cites": [ "the-mystery-stated-honestly" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-mystery-stated-honestly", "overlap": 1.0 } ], "cite_overlap_p50": 1.0, "id": "off-06", "kind": "off-page" }, { "slug": "three-librarians", "q": "how many dimensions does the embedder's map have, and what does the third librarian do?", "ok": true, "answer": "The embedder's map has 768 directions § {#the-second-librarian-searches-by-meaning}. The third librarian works only the aisle of the game being asked about, picking up the edition's own rulebook and any matched correction documents to rank passages by meaning § {#the-third-librarian-never-left-your-aisle}. The article does not state what was left out, moved, or not measured regarding the third librarian § {#the-third-librarian-never-left-your-aisle}.", "served": "The embedder's map has 768 directions § {#the-second-librarian-searches-by-meaning}. The third librarian works only the aisle of the game being asked about, picking up the edition's own rulebook and any matched correction documents to rank passages by meaning § {#the-third-librarian-never-left-your-aisle}. The article does not state what was left out, moved, or not measured regarding the third librarian § {#the-third-librarian-never-left-your-aisle}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.9510945380316116, "ttft_ms": 314, "tokens_in": 5867, "tokens_out": 117, "cached_tokens": 5568, "cites": [ "the-second-librarian-searches-by-meaning", "the-third-librarian-never-left-your-aisle" ], "gates": { "cites": [ "the-second-librarian-searches-by-meaning", "the-third-librarian-never-left-your-aisle" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "768" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-second-librarian-searches-by-meaning", "overlap": 0.75 }, { "i": 1, "cite": "the-third-librarian-never-left-your-aisle", "overlap": 0.789 }, { "i": 2, "cite": "the-third-librarian-never-left-your-aisle", "overlap": 0.375 } ], "cite_overlap_p50": 0.75, "id": "on-26", "kind": "on-page" }, { "slug": "the-new-kid", "q": "which seat did the new model pass, and which floor did it miss?", "ok": true, "answer": "The qwen3.8:27b model passed the classifier seat, clearing all five pre-registered gates § {#the-classifier-seat-five-gates-all-pass-faster}, but it missed the preservation floor of 13 in the judge seat by scoring 11 of 16 § {#the-judge-seat-and-the-floor-it-missed}. The article notes that no baseline was scored for the classifier seat and that the judge-seat rows were measured on ollama 0.32.9 while the model ran on 0.32.13 § {#the-classifier-seat-five-gates-all-pass-faster} § {#the-judge-seat-and-the-floor-it-missed}.", "served": "The qwen3.8:27b model passed the classifier seat, clearing all five pre-registered gates § {#the-classifier-seat-five-gates-all-pass-faster}, but it missed the preservation floor of 13 in the judge seat by scoring 11 of 16 § {#the-judge-seat-and-the-floor-it-missed}. The article notes that no baseline was scored for the classifier seat and that the judge-seat rows were measured on ollama 0.32.9 while the model ran on 0.32.13 § {#the-classifier-seat-five-gates-all-pass-faster} § {#the-judge-seat-and-the-floor-it-missed}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.4667854119907133, "ttft_ms": 377, "tokens_in": 15045, "tokens_out": 165, "cached_tokens": 14752, "cites": [ "the-classifier-seat-five-gates-all-pass-faster", "the-judge-seat-and-the-floor-it-missed" ], "gates": { "cites": [ "the-classifier-seat-five-gates-all-pass-faster", "the-judge-seat-and-the-floor-it-missed" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "3.8", "27", "13", "11", "16", "0.32", "9", "0.32", "13" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-classifier-seat-five-gates-all-pass-faster", "overlap": 0.611 }, { "i": 0, "cite": "the-judge-seat-and-the-floor-it-missed", "overlap": 0.778 }, { "i": 1, "cite": "the-classifier-seat-five-gates-all-pass-faster", "overlap": 0.714 }, { "i": 1, "cite": "the-judge-seat-and-the-floor-it-missed", "overlap": 0.857 } ], "cite_overlap_p50": 0.746, "id": "on-27", "kind": "on-page" }, { "slug": "the-new-kid", "q": "how does this compare to a result published in 2028?", "ok": false, "answer": "The article does not provide information regarding a result published in 2028 § {#the-arrival}.", "served": "This page doesn't say. The nearest it gets is § {#the-arrival}. The outside judges — four rival labs re-check our work. covers it more directly.", "abstained": true, "abstain_source": "gate", "abstain_section": "the-arrival", "seconds": 0.5774155160179362, "ttft_ms": 395, "tokens_in": 15045, "tokens_out": 23, "cached_tokens": 15008, "cites": [ "the-arrival" ], "gates": { "cites": [ "the-arrival" ], "cites_dropped": [], "new_noun_tokens": [ "2028" ], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": true, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [ "new noun: 1 token(s) the article does not carry" ], "new_noun_tokens": [ "2028" ], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-arrival", "overlap": 0.333 } ], "cite_overlap_p50": 0.333, "id": "off-15", "kind": "off-page" }, { "slug": "llms-txt", "q": "did any ai crawler fetch the file in the first thirty days, and how big is the largest measured llms-full file?", "ok": true, "answer": "No AI crawler asked for the file in thirty days of logs § {#does-anyone-fetch-it}, and the largest measured llms-full file weighs 30.7 MiB § {#where-the-file-came-from}. The article notes that it does not measure crawling, fetching, adoption, or agent behaviour in the wild § {#the-bench-if-the-map-arrives-does-it-help}.", "served": "No AI crawler asked for the file in thirty days of logs § {#does-anyone-fetch-it}, and the largest measured llms-full file weighs 30.7 MiB § {#where-the-file-came-from}. The article notes that it does not measure crawling, fetching, adoption, or agent behaviour in the wild § {#the-bench-if-the-map-arrives-does-it-help}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.9655687059857883, "ttft_ms": 387, "tokens_in": 12495, "tokens_out": 93, "cached_tokens": 12448, "cites": [ "does-anyone-fetch-it", "where-the-file-came-from", "the-bench-if-the-map-arrives-does-it-help" ], "gates": { "cites": [ "does-anyone-fetch-it", "where-the-file-came-from", "the-bench-if-the-map-arrives-does-it-help" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "30.7" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "does-anyone-fetch-it", "overlap": 0.385 }, { "i": 0, "cite": "where-the-file-came-from", "overlap": 0.538 }, { "i": 1, "cite": "the-bench-if-the-map-arrives-does-it-help", "overlap": 0.0 } ], "cite_overlap_p50": 0.385, "id": "on-28", "kind": "on-page" }, { "slug": "llms-txt", "q": "how many people subscribe to this site's feed?", "ok": true, "answer": "The article does not say how many people subscribe to this site's feed § {#does-anyone-fetch-it}.", "served": "This page doesn't say. The nearest it gets is § {#does-anyone-fetch-it}.", "abstained": true, "abstain_source": "model", "abstain_section": "does-anyone-fetch-it", "seconds": 0.5671258369693533, "ttft_ms": 359, "tokens_in": 12480, "tokens_out": 26, "cached_tokens": 12192, "cites": [ "does-anyone-fetch-it" ], "gates": { "cites": [ "does-anyone-fetch-it" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "does-anyone-fetch-it", "overlap": 0.167 } ], "cite_overlap_p50": 0.167, "id": "off-03", "kind": "off-page" }, { "slug": "the-open-call", "q": "how many models were handed the frozen moment, and what did the round cost?", "ok": true, "answer": "Twenty language models were handed the frozen moment § {#top}. The round's metered spend was $2.2826 § {#2-2826-over-56-metered-calls-against-38-49-of-caps}. The article notes that the agent transport holds no cost figure and no token counters § {#the-words-this-page-leans-on}.", "served": "Twenty language models were handed the frozen moment § {#top}. The round's metered spend was $2.2826 § {#2-2826-over-56-metered-calls-against-38-49-of-caps}. The article notes that the agent transport holds no cost figure and no token counters § {#the-words-this-page-leans-on}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.7220460830139928, "ttft_ms": 615, "tokens_in": 63446, "tokens_out": 88, "cached_tokens": 63168, "cites": [ "top", "2-2826-over-56-metered-calls-against-38-49-of-caps", "the-words-this-page-leans-on" ], "gates": { "cites": [ "top", "2-2826-over-56-metered-calls-against-38-49-of-caps", "the-words-this-page-leans-on" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "2.2826" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "top", "overlap": 1.0 }, { "i": 1, "cite": "2-2826-over-56-metered-calls-against-38-49-of-caps", "overlap": 0.5 }, { "i": 2, "cite": "the-words-this-page-leans-on", "overlap": 0.125 } ], "cite_overlap_p50": 0.5, "id": "on-29", "kind": "on-page" }, { "slug": "the-open-call", "q": "what did other reviewers say about this piece after it was published?", "ok": true, "answer": "The article does not state what other reviewers said about this piece after it was published § {#the-author-of-this-page-is-a-contestant-in-it}.", "served": "This page doesn't say. The nearest it gets is § {#the-author-of-this-page-is-a-contestant-in-it}. Three New Voices at the Narrator's Chair covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "the-author-of-this-page-is-a-contestant-in-it", "seconds": 1.2685149280005135, "ttft_ms": 666, "tokens_in": 63443, "tokens_out": 40, "cached_tokens": 63392, "cites": [ "the-author-of-this-page-is-a-contestant-in-it" ], "gates": { "cites": [ "the-author-of-this-page-is-a-contestant-in-it" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-author-of-this-page-is-a-contestant-in-it", "overlap": 0.429 } ], "cite_overlap_p50": 0.429, "id": "off-14", "kind": "off-page" }, { "slug": "where-your-question-goes", "q": "how long does every notebook on the server keep its entries?", "ok": true, "answer": "Every notebook on the server now keeps thirty days of entries § {#the-first-promise-your-visit-stays-between-us}. This rule applies to all notebooks, including the tour guide's, and is maintained by a nightly sweep § {#the-first-promise-your-visit-stays-between-us}. The article notes that previously, notebook rules were set to keep lines for up to a hundred years § {#the-first-promise-your-visit-stays-between-us}.", "served": "Every notebook on the server now keeps thirty days of entries § {#the-first-promise-your-visit-stays-between-us}. This rule applies to all notebooks, including the tour guide's, and is maintained by a nightly sweep § {#the-first-promise-your-visit-stays-between-us}. The article notes that previously, notebook rules were set to keep lines for up to a hundred years § {#the-first-promise-your-visit-stays-between-us}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.8624487280030735, "ttft_ms": 253, "tokens_in": 4704, "tokens_out": 106, "cached_tokens": 4416, "cites": [ "the-first-promise-your-visit-stays-between-us" ], "gates": { "cites": [ "the-first-promise-your-visit-stays-between-us" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-first-promise-your-visit-stays-between-us", "overlap": 0.875 }, { "i": 1, "cite": "the-first-promise-your-visit-stays-between-us", "overlap": 0.778 }, { "i": 2, "cite": "the-first-promise-your-visit-stays-between-us", "overlap": 0.8 } ], "cite_overlap_p50": 0.8, "id": "on-30", "kind": "on-page" }, { "slug": "where-your-question-goes", "q": "which country is the workshop's business registered in?", "ok": true, "answer": "The article does not say which country the workshop's business is registered in, though the section § {#the-first-promise-your-visit-stays-between-us} comes nearest to discussing the organization's operations.", "served": "This page doesn't say. The nearest it gets is § {#the-first-promise-your-visit-stays-between-us}.", "abstained": true, "abstain_source": "model", "abstain_section": "the-first-promise-your-visit-stays-between-us", "seconds": 0.5663619510014541, "ttft_ms": 254, "tokens_in": 4703, "tokens_out": 48, "cached_tokens": 4672, "cites": [ "the-first-promise-your-visit-stays-between-us" ], "gates": { "cites": [ "the-first-promise-your-visit-stays-between-us" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-first-promise-your-visit-stays-between-us", "overlap": 0.167 } ], "cite_overlap_p50": 0.167, "id": "off-16", "kind": "off-page" }, { "slug": "how-we-work", "q": "what does the page say about the humans and agents who make each exhibit?", "ok": true, "answer": "The page states that the human generates and judges while the agent amplifies and verifies § {#the-plain-english-version}. Agents perform tasks such as recon, planning, hardening, and drafting, while the human operator provides the initial noticing, shapes the plan, reviews the hardened version, and provides the final sign-off for release § {#the-thorough-walk}. The arrangement ensures that agents supply the receipts while judgment remains human § {#the-thorough-walk}. Regarding what was not measured, the voice trial's original five questions were never persisted § {#the-thorough-walk}.", "served": "The page states that the human generates and judges while the agent amplifies and verifies § {#the-plain-english-version}. Agents perform tasks such as recon, planning, hardening, and drafting, while the human operator provides the initial noticing, shapes the plan, reviews the hardened version, and provides the final sign-off for release § {#the-thorough-walk}. The arrangement ensures that agents supply the receipts while judgment remains human § {#the-thorough-walk}. Regarding what was not measured, the voice trial's original five questions were never persisted § {#the-thorough-walk}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.9449214730411768, "ttft_ms": 260, "tokens_in": 5164, "tokens_out": 124, "cached_tokens": 4864, "cites": [ "the-plain-english-version", "the-thorough-walk" ], "gates": { "cites": [ "the-plain-english-version", "the-thorough-walk" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-plain-english-version", "overlap": 0.375 }, { "i": 1, "cite": "the-thorough-walk", "overlap": 0.545 }, { "i": 2, "cite": "the-thorough-walk", "overlap": 0.667 }, { "i": 3, "cite": "the-thorough-walk", "overlap": 0.667 } ], "cite_overlap_p50": 0.606, "id": "on-31", "kind": "on-page" }, { "slug": "how-we-work", "q": "what is the postal address of the workshop that ran this?", "ok": true, "answer": "The article does not provide the postal address of the workshop § {#the-plain-english-version}.", "served": "This page doesn't say. The nearest it gets is § {#the-plain-english-version}.", "abstained": true, "abstain_source": "model", "abstain_section": "the-plain-english-version", "seconds": 0.4156529050087556, "ttft_ms": 316, "tokens_in": 5161, "tokens_out": 22, "cached_tokens": 5120, "cites": [ "the-plain-english-version" ], "gates": { "cites": [ "the-plain-english-version" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-plain-english-version", "overlap": 0.25 } ], "cite_overlap_p50": 0.25, "id": "off-02", "kind": "off-page" }, { "slug": "the-move", "q": "what did the median search time go from and to after the move?", "ok": true, "answer": "The median search time went from 242.0 ms in the laptop era to 77.5 ms in the new era § {#receipts}. The move put the database, the search index, and the reranker on a server-class VPS, but generation stayed on the original GPU box § {#what-moved}.", "served": "The median search time went from 242.0 ms in the laptop era to 77.5 ms in the new era § {#receipts}. The move put the database, the search index, and the reranker on a server-class VPS, but generation stayed on the original GPU box § {#what-moved}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.6926846239948645, "ttft_ms": 263, "tokens_in": 6055, "tokens_out": 69, "cached_tokens": 5760, "cites": [ "receipts", "what-moved" ], "gates": { "cites": [ "receipts", "what-moved" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "242.0", "77.5" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "receipts", "overlap": 0.9 }, { "i": 1, "cite": "what-moved", "overlap": 0.846 } ], "cite_overlap_p50": 0.873, "id": "on-32", "kind": "on-page" }, { "slug": "the-move", "q": "can you write a python script that reproduces this for me?", "ok": true, "answer": "The article does not provide a python script or instructions for writing one, though it mentions a frozen runner used for the bench § {#bench}. The text focuses on describing the migration of the production stack and the resulting performance measurements § {#what-moved}.", "served": "This page doesn't say. The nearest it gets is § {#bench}.", "abstained": true, "abstain_source": "model", "abstain_section": "bench", "seconds": 0.5988665529876016, "ttft_ms": 317, "tokens_in": 6053, "tokens_out": 51, "cached_tokens": 6016, "cites": [ "bench", "what-moved" ], "gates": { "cites": [ "bench", "what-moved" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "bench", "overlap": 0.333 }, { "i": 1, "cite": "what-moved", "overlap": 0.0 } ], "cite_overlap_p50": 0.167, "id": "off-08", "kind": "off-page" }, { "slug": "chair-trials", "q": "how many models and arms sat the five exams, and did the box stop serving?", "ok": true, "answer": "Nine models and eleven arms sat the five exams, and the box never stopped serving its real users § {#top}. The roster notes that one model ships as three builds, consisting of two quantizations and a drafter-less variant § {#roster}. The article also notes that no on-disk blob size was registered because no leg of this arc measured one § {#roster}.", "served": "Nine models and eleven arms sat the five exams, and the box never stopped serving its real users § {#top}. The roster notes that one model ships as three builds, consisting of two quantizations and a drafter-less variant § {#roster}. The article also notes that no on-disk blob size was registered because no leg of this arc measured one § {#roster}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0313718810211867, "ttft_ms": 421, "tokens_in": 30276, "tokens_out": 78, "cached_tokens": 29984, "cites": [ "top", "roster" ], "gates": { "cites": [ "top", "roster" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "top", "overlap": 0.923 }, { "i": 1, "cite": "roster", "overlap": 0.917 }, { "i": 2, "cite": "roster", "overlap": 0.909 } ], "cite_overlap_p50": 0.917, "id": "on-33", "kind": "on-page" }, { "slug": "cove-voice-head-to-head", "q": "what happened to the August class's best voice at the cove's gate?", "ok": true, "answer": "The August class's best voice was refused at the gate § {#the-verdict}. The candidate, qwen3.6:27b, failed the canon floor because it invented facts twice in thirteen prompts § {#the-verdict} and failed the latency floor because its warm median of 3 335.9 ms was slower than the bar's 2 757.6 ms § {#floor-4-latency}. No threshold table was published for the candidate because it failed its gate § {#the-verdict}.", "served": "The August class's best voice was refused at the gate § {#the-verdict}. The candidate, qwen3.6:27b, failed the canon floor because it invented facts twice in thirteen prompts § {#the-verdict} and failed the latency floor because its warm median of 3 335.9 ms was slower than the bar's 2 757.6 ms § {#floor-4-latency}. No threshold table was published for the candidate because it failed its gate § {#the-verdict}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0839675199822523, "ttft_ms": 339, "tokens_in": 16622, "tokens_out": 115, "cached_tokens": 16320, "cites": [ "the-verdict", "floor-4-latency" ], "gates": { "cites": [ "the-verdict", "floor-4-latency" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "3.6", "27", "3", "335.9", "2", "757.6" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-verdict", "overlap": 0.333 }, { "i": 1, "cite": "the-verdict", "overlap": 0.85 }, { "i": 1, "cite": "floor-4-latency", "overlap": 0.6 }, { "i": 2, "cite": "the-verdict", "overlap": 0.714 } ], "cite_overlap_p50": 0.657, "id": "on-34", "kind": "on-page" }, { "slug": "outside-judges", "q": "how many outside labs re-judged, and what was the widest cross-vendor gap?", "ok": true, "answer": "Four rival frontier labs re-judged every sealed round § {#vendors}. The widest cross-vendor gap on any arm was 0.178 § {#vendors}. This gap occurred on muse-glimmer:30b-q8_0-dflash, where mistral reads it at 0.450 against nvidia at 0.628 § {#vendors}.", "served": "Four rival frontier labs re-judged every sealed round § {#vendors}. The widest cross-vendor gap on any arm was 0.178 § {#vendors}. This gap occurred on muse-glimmer:30b-q8_0-dflash, where mistral reads it at 0.450 against nvidia at 0.628 § {#vendors}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.8841949219931848, "ttft_ms": 355, "tokens_in": 10316, "tokens_out": 82, "cached_tokens": 10016, "cites": [ "vendors" ], "gates": { "cites": [ "vendors" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "0.178", "30", "80", "0.450", "0.628" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "vendors", "overlap": 0.625 }, { "i": 1, "cite": "vendors", "overlap": 1.0 }, { "i": 2, "cite": "vendors", "overlap": 0.909 } ], "cite_overlap_p50": 0.909, "id": "on-35", "kind": "on-page" }, { "slug": "outside-judges", "q": "what is this workshop's hourly rate for consulting work?", "ok": true, "answer": "The article does not provide information regarding the workshop's hourly rate for consulting work § {#boundary}.", "served": "This page doesn't say. The nearest it gets is § {#boundary}.", "abstained": true, "abstain_source": "model", "abstain_section": "boundary", "seconds": 0.5257663140073419, "ttft_ms": 361, "tokens_in": 10309, "tokens_out": 21, "cached_tokens": 10272, "cites": [ "boundary" ], "gates": { "cites": [ "boundary" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "boundary", "overlap": 0.0 } ], "cite_overlap_p50": 0.0, "id": "off-11", "kind": "off-page" }, { "slug": "august-arrivals", "q": "how many models landed that week, and what share broke the response contract?", "ok": true, "answer": "Four open-weight models landed in one week of August 2026, and all four broke the response contract on more than 10% of calls § {#exam-one-the-judge-chair}. Specifically, the failure rates for the four candidates were 16.3%, 51.2%, 55.8%, and 62.8% § {#exam-one-the-judge-chair}. The article notes that olmo-3.1:32b-think-q4_K_M was on the seat table and in the residency census but never ran the voice exam § {#exam-two-the-narrators-chair}.", "served": "Four open-weight models landed in one week of August 2026, and all four broke the response contract on more than 10% of calls § {#exam-one-the-judge-chair}. Specifically, the failure rates for the four candidates were 16.3%, 51.2%, 55.8%, and 62.8% § {#exam-one-the-judge-chair}. The article notes that olmo-3.1:32b-think-q4_K_M was on the seat table and in the residency census but never ran the voice exam § {#exam-two-the-narrators-chair}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.1920436820364557, "ttft_ms": 324, "tokens_in": 11224, "tokens_out": 145, "cached_tokens": 10944, "cites": [ "exam-one-the-judge-chair", "exam-two-the-narrators-chair" ], "gates": { "cites": [ "exam-one-the-judge-chair", "exam-two-the-narrators-chair" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "10", "16.3", "51.2", "55.8", "62.8", "3.1", "32", "4" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "exam-one-the-judge-chair", "overlap": 0.857 }, { "i": 1, "cite": "exam-one-the-judge-chair", "overlap": 0.778 }, { "i": 2, "cite": "exam-two-the-narrators-chair", "overlap": 0.818 } ], "cite_overlap_p50": 0.818, "id": "on-36", "kind": "on-page" }, { "slug": "seat-trials", "q": "what are the kill-recall and preservation floors a model must clear?", "ok": true, "answer": "A model must achieve a kill-recall of at least 23/27 and a preservation of at least 13/16 to clear the floors § {#the-instrument}. These thresholds are written down before any model runs and each floor binds on its own § {#the-instrument}. One candidate was blocked at the billing probe and never ran § {#the-instrument}.", "served": "A model must achieve a kill-recall of at least 23/27 and a preservation of at least 13/16 to clear the floors § {#the-instrument}. These thresholds are written down before any model runs and each floor binds on its own § {#the-instrument}. One candidate was blocked at the billing probe and never ran § {#the-instrument}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.7427470050170086, "ttft_ms": 288, "tokens_in": 7192, "tokens_out": 79, "cached_tokens": 7168, "cites": [ "the-instrument" ], "gates": { "cites": [ "the-instrument" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "23/27", "13/16" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-instrument", "overlap": 0.6 }, { "i": 1, "cite": "the-instrument", "overlap": 1.0 }, { "i": 2, "cite": "the-instrument", "overlap": 1.0 } ], "cite_overlap_p50": 1.0, "id": "on-37", "kind": "on-page" }, { "slug": "seat-trials", "q": "how many models will sit this exam next year?", "ok": true, "answer": "The article does not say how many models will sit this exam next year § {#the-instrument}.", "served": "This page doesn't say. The nearest it gets is § {#the-instrument}. The Instrument Travels covers it more directly.", "abstained": true, "abstain_source": "model", "abstain_section": "the-instrument", "seconds": 0.4151677449699491, "ttft_ms": 273, "tokens_in": 7188, "tokens_out": 21, "cached_tokens": 6912, "cites": [ "the-instrument" ], "gates": { "cites": [ "the-instrument" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-instrument", "overlap": 0.143 } ], "cite_overlap_p50": 0.143, "id": "off-21", "kind": "off-page" }, { "slug": "voice-trials", "q": "how many model tags were tried, and what does the anchor model read?", "ok": true, "answer": "The article states that twenty-four distinct model tags were evaluated across forty scored runs § {#provenance-4cae43}. The anchor model reads 9.0, 7.0, and 6.2 on three different record sets § {#the-method-limits-first}. As a note, the article mentions that no delta across record sets is ever published and that the 2026-08 field never sampled a per-arm resident size § {#the-2026-08-field-six-arms-on-the-96g-workstation-blind}.", "served": "The article states that twenty-four distinct model tags were evaluated across forty scored runs § {#provenance-4cae43}. The anchor model reads 9.0, 7.0, and 6.2 on three different record sets § {#the-method-limits-first}. As a note, the article mentions that no delta across record sets is ever published and that the 2026-08 field never sampled a per-arm resident size § {#the-2026-08-field-six-arms-on-the-96g-workstation-blind}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.0740319699980319, "ttft_ms": 307, "tokens_in": 10338, "tokens_out": 129, "cached_tokens": 10048, "cites": [ "provenance-4cae43", "the-method-limits-first", "the-2026-08-field-six-arms-on-the-96g-workstation-blind" ], "gates": { "cites": [ "provenance-4cae43", "the-method-limits-first", "the-2026-08-field-six-arms-on-the-96g-workstation-blind" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "9.0", "7.0", "6.2", "2026-08" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "provenance-4cae43", "overlap": 0.6 }, { "i": 1, "cite": "the-method-limits-first", "overlap": 1.0 }, { "i": 2, "cite": "the-2026-08-field-six-arms-on-the-96g-workstation-blind", "overlap": 0.6 } ], "cite_overlap_p50": 0.6, "id": "on-38", "kind": "on-page" }, { "slug": "licences", "q": "what does the page say it does with each model's licence?", "ok": true, "answer": "The page states that it reads every model's licence first-hand and stamps it with the date it was last read in full § {#how-to-read-this-page}. These reads are performed using specific methods named per table, such as reading the licence file packaged with the weights or fetching permission flags from a platform's raw record § {#the-painters-diffusion-read-2026-08-11-except-where-a-row-says-otherwise}. The article notes that unread models are withheld and that some models, such as gemma2:9b, are not currently installed on the bench box and thus cannot be re-read § {#provenance-063a00}.", "served": "The page states that it reads every model's licence first-hand and stamps it with the date it was last read in full § {#how-to-read-this-page}. These reads are performed using specific methods named per table, such as reading the licence file packaged with the weights or fetching permission flags from a platform's raw record § {#the-painters-diffusion-read-2026-08-11-except-where-a-row-says-otherwise}. The article notes that unread models are withheld and that some models, such as gemma2:9b, are not currently installed on the bench box and thus cannot be re-read § {#provenance-063a00}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.1413326140027493, "ttft_ms": 262, "tokens_in": 8549, "tokens_out": 155, "cached_tokens": 8256, "cites": [ "how-to-read-this-page", "provenance-063a00" ], "gates": { "cites": [ "how-to-read-this-page", "provenance-063a00" ], "cites_dropped": [ "the-painters-diffusion-read-2026-08-11-except-where-a-row-says-otherwise" ], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [ "cite: 1 cite(s) not in this page's anchors: the-painters-diffusion-read-2026-08-11-except-where-a-row-says-otherwise" ], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [ "the-painters-diffusion-read-2026-08-11-except-where-a-row-says-otherwise" ] }, "figures": [ "2", "9" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "how-to-read-this-page", "overlap": 0.636 }, { "i": 2, "cite": "provenance-063a00", "overlap": 0.2 } ], "cite_overlap_p50": 0.418, "id": "on-39", "kind": "on-page" }, { "slug": "hear-it-for-yourself", "q": "how many clips does the page carry, and how many presentations made up the sitting?", "ok": true, "answer": "The page carries eighty-four clips § {#the-twelve-requests} and the sitting was composed of thirty-six presentations § {#the-blind-sheet-unsealed}. These presentations consisted of twenty-four scoring pairs and twelve sentinels § {#the-blind-sheet-unsealed}. The article notes that the run actually rendered eighty-five clips, but the extra clip was a determinism control § {#how-to-check-these-clips}.", "served": "The page carries eighty-four clips § {#the-twelve-requests} and the sitting was composed of thirty-six presentations § {#the-blind-sheet-unsealed}. These presentations consisted of twenty-four scoring pairs and twelve sentinels § {#the-blind-sheet-unsealed}. The article notes that the run actually rendered eighty-five clips, but the extra clip was a determinism control § {#how-to-check-these-clips}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 0.9897119760280475, "ttft_ms": 333, "tokens_in": 19025, "tokens_out": 97, "cached_tokens": 18720, "cites": [ "the-twelve-requests", "the-blind-sheet-unsealed", "how-to-check-these-clips" ], "gates": { "cites": [ "the-twelve-requests", "the-blind-sheet-unsealed", "how-to-check-these-clips" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-twelve-requests", "overlap": 0.429 }, { "i": 0, "cite": "the-blind-sheet-unsealed", "overlap": 0.714 }, { "i": 1, "cite": "the-blind-sheet-unsealed", "overlap": 0.857 }, { "i": 2, "cite": "how-to-check-these-clips", "overlap": 0.7 } ], "cite_overlap_p50": 0.707, "id": "on-40", "kind": "on-page" }, { "slug": "the-ceiling-is-not-the-corpus", "q": "how did the adapter score against the base model in the listening, and on how many house prompts did it win?", "ok": true, "answer": "The adapter did not clear the listening test, as the listener preferred the untouched model on 11 of the 24 pairs and the adapter on 4 § {#the-blind-sitting}. On the four requests that specifically asked for house, the adapter won zero times, with the untouched model winning every pair the listener could separate § {#what-the-adapter-actually-did}. The article notes that the name tag alone moved renders toward the corpus in 3 of 12 cases, while the weights moved them in 7 of 12 § {#what-the-adapter-actually-did}.", "served": "The adapter did not clear the listening test, as the listener preferred the untouched model on 11 of the 24 pairs and the adapter on 4 § {#the-blind-sitting}. On the four requests that specifically asked for house, the adapter won zero times, with the untouched model winning every pair the listener could separate § {#what-the-adapter-actually-did}. The article notes that the name tag alone moved renders toward the corpus in 3 of 12 cases, while the weights moved them in 7 of 12 § {#what-the-adapter-actually-did}.", "abstained": false, "abstain_source": null, "abstain_section": null, "seconds": 1.121516026032623, "ttft_ms": 386, "tokens_in": 9221, "tokens_out": 126, "cached_tokens": 8928, "cites": [ "the-blind-sitting", "what-the-adapter-actually-did" ], "gates": { "cites": [ "the-blind-sitting", "what-the-adapter-actually-did" ], "cites_dropped": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "redacted": false, "figure_unverified": false, "uncited": false, "new_noun": false, "blocked": false, "section_marker": false, "directive_hit": false }, "gate_detail": { "reasons": [], "new_noun_tokens": [], "new_noun_tokens_withheld": 0, "cites_dropped": [] }, "figures": [ "11", "24", "4", "3", "12", "7", "12" ], "figures_verbatim": true, "cites_valid": true, "redaction_hits": 0, "cite_overlap": [ { "i": 0, "cite": "the-blind-sitting", "overlap": 1.0 }, { "i": 1, "cite": "what-the-adapter-actually-did", "overlap": 0.688 }, { "i": 2, "cite": "what-the-adapter-actually-did", "overlap": 0.923 } ], "cite_overlap_p50": 0.923, "id": "on-41", "kind": "on-page" } ], "verdicts": { "on_page_answered": true, "non_verbatim_figures": true, "off_page_abstained": true, "redaction_hits": true, "p50_warm_seconds": true, "p95_warm_seconds": true }, "passed": true }