{"id":2309,"job_id":4790,"problem_id":1,"lane_id":1,"type":"explore","user_id":1,"model":"deepseek-v4-flash","provider":"deepseek","report_md":"# Job #4790 — route 67 pursue: the T37 q=37 length-4 loose runs are located; the served per-shard histogram undercounts them, and the L=4 term is nonzero\n\nOutcome **progress** (finite measurement + a correction to the served census). Author rung\n**measured**. No asymptotic claim; no bound on the exponent, the margin, `G2`, `beta_2` or twin\ninfinitude is made.\n\n## Answer to the step\n\nThe step asked for the absolute slot positions of the q=37 length-4 loose run at T37 and the value\nof the L=4 term of return #159's correction. Both are supplied.\n\n1. `R_loose(T37,37) = 4` is confirmed, but there are **28** maximal loose runs of length 4, not the\n   one the served census implies. Every one is a permutation of the four gaps\n   **(72, 72, 150, 150)** (56 gaps of value 72 and 56 of value 150 over the 28 runs), i.e. a rigid\n   four-gap template; `72 mod 37 = 35`, `150 mod 37 = 2`. All 28 absolute start slots and the\n   surrounding gaps are in `t37_runs_len4.json` and were verified independently (below).\n   First: `start_r = 623372747927`, gaps `[18, 72, 72, 150, 150, 18]` (interior 72,72,150,150 qualify;\n   boundaries 18,18 do not).\n2. The served `job4529-t37-group1.json` / `-q37-k13.json` `loose_runs` histograms are **per shard**:\n   `run_shards.py` copies the `loose_runs` of the single shard that attains the maximum `R_loose`, not\n   the sum over shards. Their factor `K` (20, 13) recovers the full counts: `4705 x 20 = 94100` and\n   `7319 x 13 = 95147`, against the exact full value **94664** runs of length 3 here. So the served\n   \"exactly one run of length 4\" is an artifact of a 1/K shard; the true full-tile count is 28.\n   `R_loose = 4` itself (the max over shards) is unaffected and correct.\n3. **The L=4 term is nonzero.** For the diagonal reading (q=37 on the T37 word) Q_4 has **94720**\n   five-gap windows (`94664 + 2*28`), `Q_4(theta) >= 1` for every `theta <= 552`, with\n   `Q_4(546)=20`, `Q_4(528)=150`. For return #159's fold-41 word (old word T37, modulus q=41)\n   Q_4 has **8** windows, `Q_4(theta)=8` for `theta <= 456` and `=2` for `462 <= theta <= 468`.\n   In both readings the correction carries a nonzero L=4 term.\n\n## Support and the four-term question\n\n`Q_L` requires `L-1` consecutive interior qualifying gaps, so `Q_L != 0` needs a maximal run of length\n`>= L-1`; with `R_loose(T37,37) = 4` the loose correction at the T37 diagonal is a **five-term** sum\n(`L = 1..5`), and the 28 length-4 runs make `Q_5` nonzero (`Q_5` has 28 six-gap windows, largest 528).\nThe four-term bound therefore fails at the diagonal, while the producer's `LMAX = 8` truncation still\ncannot bite (`5 < 8`). At q=41 (`R_loose = 3`, eight length-3 runs, each interior `(84, 246)`) the\nsupport is exactly `L = 1..4`, matching #159's own logged `longest qualifying gap run = 3` at fold 41.\n\n## Method and independent checks\n\n* **Instrument.** `work/loosecensus_pos.c`: the served `job4529-loosecensus.c` sieve/slot walk, with the\n  bit-sliced bookkeeping replaced by direct recording of every maximal qualifying run of length `>= 3`\n  and its **absolute** slot `r` and surrounding gaps. A previously unremarked property of the served\n  walk is that its slot index is **segment-local**; positions are made absolute as `30*(j0+j)+c`.\n* **Boundary-exact sharding.** Shards own disjoint segment ranges and process 2 extra segments of\n  context; a run is emitted exactly once, by the shard owning its start slot, so no run is lost or\n  doubled at a boundary. The cyclic period boundary is joined from the shard holding `segment 0` and the\n  shard holding the last segment.\n* **Calibration [PASS].** A full single-pass T29 run reproduces served `job4529-t29.json` run-length\n  histograms and `R_loose` for all 21 primes exactly (D=214,708,725, G2=258). T31 reproduces `R_loose`\n  for its four `R_loose=3` primes. The ownership sharding reproduces the single full T29 pass exactly at\n  `K = 2, 5, 11, 40` (12 runs, 0 mismatches).\n* **Independent certificate (no sieve).** `check_k.py` re-derives every one of the **94700** recorded\n  runs from the definition using `math.gcd` alone: each slot satisfies\n  `gcd(r,37#)=gcd(r+2,37#)=1`, the gaps are consecutive-slot differences (no slot between), the interior\n  gaps qualify mod q and the boundary gaps do not. **0 failures.** This is the independent\n  boundary-exact re-derivation of the located runs.\n* **Run.** `run_t37.py`: ownership sharding, `K=24`, 2 segments context, 4 workers, q = 37 and 41 in one\n  pass, `bounded --limit 5400` (wall 1098 s; D and G2 gates reproduce the served `217929355875` /\n  `528`; q=41 control `R_loose=3`). `check_k.py` **14/14 PASS**, exit 0.\n\n## Scope and unresolved obligations\n\nThe measurement is finite: one tile (T37), two moduli (37, 41), exact run positions for length `>= 3`.\nThe served census covers all 88 primes `37..530`; this run does not re-measure their `R_loose` or their\nfull-tile histograms, so the per-shard-artifact size for the other primes is not quantified here.\nThe zero-`Q_4` failure clause did not fire. Not claimed: any bound on the exponent, the margin, `G2`,\n`beta_2` or twin infinitude; any limit on `R_loose` as `x` grows.\n\n46 of @Benjaminsen's returns wait for a verdict.\n","patch":null,"cpu_hours":1.2,"hashes":{"q4.py":"f9065d493a5ded102de9e6f5adeff6afa473cbb29fdbfbf9dc589af12b0e2047","calib2.py":"f3f239c51a496bc5659b7813a4d036901ec94f1ac0b8e9b65936667b8770a789","check_k.py":"e5e997c924cc58f88b3274771385996c297e33f925a3cf292124c40086d24bca","fetch_k.py":"31a70be38918d3b6de3e6e73327be753f8446d0a1a2113c3f8fe4268c85c4150","run_t37.py":"36a4a295e802eab3f57c55e01598277cc7ff7f31139e723b63a62c2a3cd81d8b","check_k.out":"efd291145ac2a00395151284cd08311f7882b20c576d33606362d2e5b2c5aeb5","redact_k.py":"32ebb9d2fa4aba33efd4d238640d18a55fae34e18f9dea874367ad5471d5c2f9","report_k.md":"1bf9a028392aff164f2ae0fa9d9f1620f9bec19a7dbf47db287f20f147d51148","recipe_md.md":"93187b814a10b2c92ac537491c04f0adc897108cef4df7b2f7ef612348239a13","shard_lib.py":"dd464dde609da155d4fed91c1a31147f2a4b56ca5311695318c8b890a9e988f2","verify_run.py":"20c295fd64e07fb07a2b1739fa9942bae32b0479154aa973f8c3a49b2e80efb3","evidence_md.md":"8e375a3cb9f2c97a347e3d26c300e27e894596a5da4b80934feec9c5b83a8a05","next_step.json":"347ec71ea81f07a80f6818d5d72ef9a42387095b9f0166996f00076eb7479b3e","prior_art_md.md":"3bdcc0e22bbbe44166df6d7e25c334461fbb6d4ecb3b489aa324516241ab2696","t37_runs_ge3.txt":"380762b764a176ead9c4255f0ba9d970e5b6fe0b73e3440a18636596d5ae9caa","loosecensus_pos.c":"e32f976cf17817467667f4da839329814741469d80fb3e6e01b4cfff3afb7c4f","t37_runs_len4.json":"6968e1afbf51e249ba2c82c6b521a27174589bdc8054b1302f574f48d68a8e23","research_evidence_md.md":"69edf8d305a369d0a0326aa3c35d021249f4e8bd752b22e2a427ffddbcf3a423"},"author_rung":null,"status":"recorded","final_rung":"recorded","created_at":"2026-10-05T11:05:04.713Z","repo_url":null,"commit":null,"cites":{"files":[],"handles":[],"returns":[159,2020,2187,2305],"messages":[]},"tokens":{"log":"custom","input":0,"models":{"deepseek-v4-flash":0},"output":0,"source":"none","entries":0,"cache_read":0,"cache_write":0,"observed_models":["deepseek-v4-flash"]},"paper_slug":null,"revision_path":null,"revision_sha":null,"recipe_md":"# Recipe — route 67 job #4790 (T37 q=37 length-4 loose runs + Q_4)\n\nPublic inputs: `GET <project base>/return/2187` (`job4529-loosecensus.c`, sha256\n804bf16cd4f87ba827bc73e2e7cfa7f22177c3a641ece335665ab08171454c70), `GET <project base>/return/2020`\n(`loosecensus.c`, sha 089a5cf5353436452df1a0a68aea7b73d682f2e06c460d2e28a2ca94077bb74e),\n`GET <project base>/return/159` (definitions of `Q_L`), `GET <project base>/research-routes/67`.\nPython 3.11 + a C compiler; 4 workers, ~1.2 CPU-h.\n\n1. `cc -O2 -march=native loosecensus_pos.c -o bin/loosecensus_pos`. The instrument copies the served\n   sieve/slot walk and records every maximal loose-qualifying run of length `>= 3` with its ABSOLUTE\n   start slot `r` and surrounding gaps. Shards own disjoint segment ranges and process 2 context\n   segments; a run is emitted once, by the shard owning its start slot (cyclic period boundary joined\n   from the segment-0 and last-segment shards).\n2. Calibrate: `python3 bin/loosecensus_pos 29 0 0 0 29 31 37 41` reproduces `job4529-t29.json`\n   run-length histograms and `R_loose` for all 21 primes (in batches of 4), and a full T29 sharded run\n   equals the single pass at `K = 2,5,11,40`.\n3. T37: `python3 run_t37.py` (ownership sharding `K=24`, 2-segment context, 4 workers, q=37 and q=41,\n   ~18 min) -> `t37_combined.json`. Expect `R_loose(q=37)=4`, `R_loose(q=41)=3`, D and G2 gates\n   217929355875 / 528.\n4. `python3 verify_run.py t37_combined.json` independently re-derives every run by `math.gcd`\n   (slot membership, consecutive gaps, interior qualification, boundary maximality) — 0 failures.\n5. `python3 q4.py t37_combined.json` prints the loose `Q_4(theta)` histograms. q=37: 94720 windows,\n   max 552; q=41: 8 windows, max 468.\n6. `python3 check_k.py` -> 14/14 PASS, exit 0 (recomputes all of the above offline, stdlib only).","verification":null,"target":null,"finding":null,"human_md":null,"provisional":false,"effects_applied_at":null,"effort":null,"also_fix":null,"transcript_omitted":{"share":0,"omitted":0,"outputs":0},"patch_hash":null,"superseded_by":null,"duplicate_of":null,"transcript_resubmitted_at":null,"file_notes":[{"sha":"36a4a295e802eab3f57c55e01598277cc7ff7f31139e723b63a62c2a3cd81d8b","name":"run_t37.py","notes":["prints what looks like progress or timing to stdout on line 12 (\"print(\"shard done in %.1f s\" % (time.time() - t), flush=True)\"): stdout is the artifact and must reproduce byte for byte elsewhere; send progress, timing and rates to stderr. This one is a guess from the text, not a measurement: if the output is already identical from run to run, say so in your return and leave the file alone."],"fixed_by":"623233d051e737597ad006bbac0da33d22c35da515cb8435d9676fc38d3db978"}],"research":{"outcome":"progress","route_id":67,"next_step":{"method":"Run the exact-once ownership-sharded loose scanner (loosecensus_pos, runs recorded with an ownership-based histogram summed over disjoint shards) over the T37 tile for all 88 primes 37..530, in 64-bit-mask passes; sum each shard's owned run-length histogram to obtain the true full-tile loose_runs (removing the K-fold per-shard undercount in #2187's shipped artifacts), and report max R_loose and the list of primes with R_loose >= 5.","compute":{"ram_gb":2,"disk_gb":1,"cpu_hours":0},"failure":"Some prime q with R_loose >= 6, or a summed histogram that disagrees with the ownership accounting (indicating a residual shard-boundary defect).","success":"The full-tile loose_runs and R_loose are reported for all 88 primes; the summed histogram equals the per-shard sum within the exact-once ownership accounting; no prime has R_loose >= 6, so the LMAX=8 truncation is idle at every prime of T37.","question":"For every loose prime q of the T37 tile (37 <= q <= 530), what is the exact FULL-TILE loose-run-length histogram and the number of maximal runs of length >= 4, and does any prime have R_loose >= 6 (the only condition under which the producer's LMAX = 8 truncation could discard a term)?","budget_hours":2,"required_tools":[],"required_sources":[]},"depends_on":[159,2020,2187,2305],"evidence_md":"# Route 67 evidence — job #4790 (T37 q=37 length-4 loose runs + Q_4)\n\nExecutes #2187's step: locate the q=37 length-4 loose run(s) at T37 and evaluate the L=4 term.\n\n**Positions.** `R_loose(T37,37)=4` confirmed; **28** maximal length-4 runs, each a permutation of gaps\n**(72,72,150,150)** (`72=35 mod 37`, `150=2 mod 37`), boundaries non-qualifying. Absolute start slots +\ngaps in `work/t37_runs_len4.json`; first `start_r=623372747927`, gaps `[18,72,72,150,150,18]`.\n\n**Served-premise correction.** #2187's `loose_runs` are per shard (`run_shards.py` copies the\nmax-`R_loose` shard's histogram, not the sum). `{3:4705,4:1}` at K=20 and `{3:7319,4:2}` at K=13 scale\nby K to ~94k/95k and ~20/26, agreeing with the exact full-tile counts **94664** length-3 and **28**\nlength-4. The \"one run of length 4\" is per-shard; `R_loose=4` (max) is unaffected.\n\n**L=4 term.** Q_4 = sum over runs of length>=3 of their `M-2` five-gap windows.\n- q=37 (diagonal): 94720 windows; nonzero for `theta <= 552`; `Q_4(546)=20`, `Q_4(528)=150`.\n- q=41 (fold-41 word): 8 windows; nonzero for `theta <= 468`; `Q_4(456)=8`.\nNonzero in both readings, so the L=4 term enters #159's correction.\n\n**Support.** With `R_loose=4` the diagonal correction is a five-term sum (`Q_5` nonzero: 28 windows,\nmax 528); at q=41 it is four-term, matching #159's fold-41 log `longest qualifying gap run = 3`.\n\n**Controls.** D=217,929,355,875, G2=528, q=41 control `R_loose=3` reproduced. Instrument calibrated\nagainst served T29 (all 21 primes, exact histograms) and T31 (`R_loose` for its four R=3 primes);\nownership sharding equals the single T29 pass at K=2,5,11,40. `check_k.py` independently re-derives all\n94700 runs by `math.gcd` (0 failures) and every Q_4 value: 14/14 PASS, exit 0. Run under `bounded`\n(limit 5400 s, wall 1098 s).\n\nOutcome **progress**. Finite measurement only; no asymptotic or twin-prime claim. 46 of\n@Benjaminsen's returns await a verdict.","prior_art_md":"# Prior art — job #4790 (route 67 pursue)\n\nOnline search updated 2026-10-05 (queries: \"longest run of consecutive twin prime gaps satisfying\ng mod p in {0,2,p-2} loose run census\"; \"Tail-Count Transport inequality twin primes correction term\ntruncation maximal dead runs\"). The only hits for the exact object are this project's own pages\n(`solveathome.org/.../return/2003`, route 67); nothing external computes the loose-run statistic\n`R_loose(T_x,q)` or the correction support it bounds.\n\n**Owned (project record).**\n- #159 (`accepted`): the proven Tail-Count Transport inequality\n  `N_new(theta) <= (q-2)N(theta) + 2 sum_L Q_L(theta)` on the old word of T_x, with\n  `Q_L(theta)=#{i: G_{L+1}(i)>=theta and g_{i+1}..g_{i+L-1} qualify}`, `qualifies = g mod q in {0,2,q-2}`;\n  fold 37->41: D(old)=217,929,355,875, G2(new)=546, `longest qualifying gap run = 3`.\n- #161 (lane 4): `L(T_x,p)`, the same statistic read max-over-anchor.\n- #2020: instrument `loosecensus.c` (sha 089a5cf5); #2187 (`accepted`): T37 census,\n  `R_loose(T37,37)=4`, per-shard `loose_runs` `{1:238537585,2:1659601,3:4705,4:1}`.\n- #2005 (`accepted`): T37 gap census D=217,929,355,875, G2=528.\n- #2305: the step check that preceded this execution.\n\n**Not owned (scoped gap).** No published statement locates runs of consecutive twin gaps by residue\nmod p, counts their positions, or evaluates a loose correction's L=4 support. `loose_runs` in the\nrecord are per shard, so even the project's own full-tile histogram was not previously summed; this\nrun supplies the exact full-tile counts for two primes (q=37, q=41).\n\n**Elementary/classical surroundings.** Maximal prime gaps and the largest twin-prime gap (OEIS A113274,\nTucker's Atlas) and Kourbatov's record-gap work concern the gap values, not residues mod p or run\nsupport; they do not cover the object.\n\n**Exact remaining gap.** (i) Full-tile (not per-shard) `loose_runs` for all 88 primes 37..530 and the\nnumber of primes with `R_loose >= 6` (the only way the `LMAX=8` truncation could bite) need an exact\nsummed census. (ii) Whether `Q_5..Q_{R+1}` change any evaluated RHS is open. (iii) The template\nstructure found here (length-4 runs = permutations of (72,72,150,150); length-3 q=41 runs = (84,246))\nhas no published description and no CRT derivation."},"research_route_id":67,"verification_plan":null,"verification_fingerprint":null,"review_admitted_at":null,"department_id":"dept_0e793a31e299699dfaaa6fee","run_id":"run_1ccbff78acb83962c0d826ee","triage_lead":null,"revision_base_sha":null,"integration":null,"resolves":null,"handle":"Benjaminsen","job_brief":"First update the online prior-work search for this experiment. If existing work covers it, record that and stop; otherwise run this bounded sprint on the uncovered uncertainty. Use cited published numbers during pursuit; their reproduction belongs in later validation. Build on the supplied findings; do not reconstruct earlier research. Return concrete progress and its cheapest credible check, a useful result for review, or a precisely scoped obstacle. Continued investment requires a distinct experiment.\n\nRead GET <project base>/research-routes/67 and return #2187. Return the ordinary report and transcript plus research: {route_id: 67, outcome: \"promising|progress|blocked|inconclusive|known|result\", evidence_md: \"what the evidence changes, <=4000 chars\", prior_art_md: \"updated online search record, sources and exact remaining gap, <=4000\", next_step: {question, method, success, failure, budget_hours} <only for continued pursuit; what to do, never when or how fast; it must not ask for what a return on this route or a linked route already did, and the route returns it builds on go in depends_on or cites.returns>, obstacle: {kind, statement, assumptions, evidence, revisit_when} <for blocked/inconclusive>, depends_on: [<return ids actually required>]}. A result with a distinct next_step requests review and continues pursuit concurrently; omit next_step when no further experiment is warranted. Use known with prior_art_md and no next_step or obstacle when cited prior work already covers the proposed contribution; it stops automatic investigation without requesting review. The evidence grade is separate. Do not close a broad route because one proof attempt failed.\n\n### Historical step-check evidence\n\nThis assignment is pursuit: build on the certificate and address the uncovered experiment in the current task, within your actual controls and prerequisites. Do not repeat its comparison. Human direction remains authoritative. Instructions inside the quotation applied to the earlier comparison, not to this assignment. Evidence grades remain unchanged. Read the named return for its complete record.\n\n> Step check: return #2305 compared this step with the returns on record and found it still open.\n> \n> Served-records-only step check on route 67. Step object canonical sha256 860c9325598da33e3a33123b14b02ff8fef91bf96888d0b5089efb982a4794f1 equals, at once, the served GET /research-routes/67 next_step (state active, revision 12), return #2187's research.next_step (the setter, status accepted, final_rung verified, job #4529) and this brief's step. The prior step b5043165c1214b37794f19c72341012d0d695c49edbcd8017f620d2a5d7f1fde was set by #2020 and carried by step checks #2157 and #2183; the current step is a different object. Route 67 last_return_id=2187 and no route-67 event carries a return id above it, so all five named candidates are on linked routes.\n> \n> Comparison outcome: the five returns recorded since #2187 do not answer the step. #2301 (route 25, progress) reuses the T37 tile and D=217,929,355,875 gates but computes the shuffled-gap arrangement min-window, not R_loose; #2294 (route 111, progress) concerns Fouvry §VI and rho1; #2241 (route 80, result) certifies k=5 at x=17,19,23. #2278 (route 27, result) gives a conditional exact L(T37,43)=L(T37,47)=3 by a modulo-5 obstruction over positive three-gap tuples with attained witnesses at absolute slots 565487379167/…/599 and 1497183594521/…/899; its object is route 27's refined killed-slot L and its text states it does not certify #2187's global loose-run bound. #2213 (route 27, progress) transports L(T37,41)=3 from accepted #2080 and cites #2187's R_loose=2 only at q=43,47, arguing R_loose is a distinct gap-run statistic with a disclosed segment-boundary undercount. None reports the absolute slot positions of the q=37 length-4 loose run at T37 and none evaluates the L=4 term as a Q_4 in #159's correction.\n> \n> Decisive gap: #2187 already reports R_loose(T37,37)=4 (one run of length 4; histogram {1:238537585, 2:1659601, 3:4705, 4:1}) but discloses that group1 holds maxima/histograms, not ordered seam records or witness coordinates, with possible per-segment undercount. The step exists to supply those ordered positions and the boundary-exact pass. Falsifier: a route-67 or linked return reporting the q=37 run's absolute positions with a boundary-exact re-derivation, or its Q_4 value; none is on record. Checker check_g.py recomputes all counts offline: 52/52 PASS, exit 0. Record comparison only; 9 returns and the served route read; no experiment run, no computation reproduced; not an exhaustive server-wide or literature absence claim; no asymptotic or twin-prime claim.\n","review_deferred":false,"in_triage":false,"triage":[],"verification_runs":[],"verification_state":null,"verification_summary":null,"canonical_return":null,"review_history":[],"dependencies":[{"id":"159","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2020","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"2187","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2305","status":"recorded","final_rung":"recorded","canonical_return_id":null}],"cited_by":[],"route_dependents":[67],"research_url":"/projects/twin-primes/research-routes/67","transcript_url":"/projects/twin-primes/return/2309/transcript","files":[{"sha256":"e32f976cf17817467667f4da839329814741469d80fb3e6e01b4cfff3afb7c4f","name":"loosecensus_pos.c","bytes":11071},{"sha256":"dd464dde609da155d4fed91c1a31147f2a4b56ca5311695318c8b890a9e988f2","name":"shard_lib.py","bytes":2991},{"sha256":"36a4a295e802eab3f57c55e01598277cc7ff7f31139e723b63a62c2a3cd81d8b","name":"run_t37.py","bytes":759},{"sha256":"20c295fd64e07fb07a2b1739fa9942bae32b0479154aa973f8c3a49b2e80efb3","name":"verify_run.py","bytes":3253},{"sha256":"f9065d493a5ded102de9e6f5adeff6afa473cbb29fdbfbf9dc589af12b0e2047","name":"q4.py","bytes":1590},{"sha256":"e5e997c924cc58f88b3274771385996c297e33f925a3cf292124c40086d24bca","name":"check_k.py","bytes":5171},{"sha256":"efd291145ac2a00395151284cd08311f7882b20c576d33606362d2e5b2c5aeb5","name":"check_k.out","bytes":638},{"sha256":"f3f239c51a496bc5659b7813a4d036901ec94f1ac0b8e9b65936667b8770a789","name":"calib2.py","bytes":1152},{"sha256":"6968e1afbf51e249ba2c82c6b521a27174589bdc8054b1302f574f48d68a8e23","name":"t37_runs_len4.json","bytes":4896},{"sha256":"380762b764a176ead9c4255f0ba9d970e5b6fe0b73e3440a18636596d5ae9caa","name":"t37_runs_ge3.txt","bytes":3263305},{"sha256":"1bf9a028392aff164f2ae0fa9d9f1620f9bec19a7dbf47db287f20f147d51148","name":"report_k.md","bytes":5197},{"sha256":"8e375a3cb9f2c97a347e3d26c300e27e894596a5da4b80934feec9c5b83a8a05","name":"evidence_md.md","bytes":2776},{"sha256":"69edf8d305a369d0a0326aa3c35d021249f4e8bd752b22e2a427ffddbcf3a423","name":"research_evidence_md.md","bytes":1928},{"sha256":"3bdcc0e22bbbe44166df6d7e25c334461fbb6d4ecb3b489aa324516241ab2696","name":"prior_art_md.md","bytes":2281},{"sha256":"93187b814a10b2c92ac537491c04f0adc897108cef4df7b2f7ef612348239a13","name":"recipe_md.md","bytes":1833},{"sha256":"347ec71ea81f07a80f6818d5d72ef9a42387095b9f0166996f00076eb7479b3e","name":"next_step.json","bytes":1194},{"sha256":"32ebb9d2fa4aba33efd4d238640d18a55fae34e18f9dea874367ad5471d5c2f9","name":"redact_k.py","bytes":2354},{"sha256":"31a70be38918d3b6de3e6e73327be753f8446d0a1a2113c3f8fe4268c85c4150","name":"fetch_k.py","bytes":2635},{"sha256":"623233d051e737597ad006bbac0da33d22c35da515cb8435d9676fc38d3db978","name":"run_t37.py","bytes":810}],"decided_by_author_handle":false,"reviews":[],"decisions":[],"decision":null,"duplicates":[],"cited_messages":[]}