Paper shape
real
Question, methods, Phase 0 evidence, Phase 1A freeze, completed runs, and claim boundaries are documented.
Submission state
ready
Source release and DOI are complete; workshop/artifact and preprint packaging can proceed with explicit biological limitations.
Main blocker
review
The experimental unit and artifact archive are complete; biological review and venue polish remain.
Done
| Area | Artifact | Status |
|---|---|---|
| Manuscript skeleton | paper/v1_workshop.md | drafted |
| Public manuscript page | workshop draft | published |
| Pandoc-rendered draft | pandoc draft | published |
| Figure 1 architecture | pipeline_architecture.svg | added to manuscript |
| Phase 0 result packet | public tunnel smoke | published |
| Phase 1 protocol | protocol page | published |
| Target candidates | CSV / JSON | 24/24 fetch-checked |
| Model matrix | model_matrix_v1.csv | published |
| Runner | PDB:CHAIN target labels | implemented |
| PDB fetch Worker | auth-chain matching | deployed |
| 3-target rehearsal | phase1-rehearsal-3target | technical run complete |
| Direct long-target route | phase1-direct-long-13bb | technical recovery complete |
| SimpleFold-3B smoke | phase1-simplefold3b-short | install smoke complete |
| SimpleFold-3B mid CLI smoke | phase1-simplefold3b-mid-13cz-cli | orchestration smoke complete |
| SimpleFold-3B telemetry smoke | phase1-simplefold3b-mid-13cz-cli-telemetry | peak RSS captured |
| SimpleFold-3B execution metadata smoke | phase1-simplefold3b-mid-13cz-cli-execmeta | execution metadata captured |
| Repeat/cache protocol | repeat-cache-protocol | published |
| No-fixture manual Reader bridge | phase1-no-fixture-manual-reader | manual bridge complete |
| Target review triage | target-review-triage | 8 keep / 14 review / 2 replace |
| SimpleFold-3B benchmark slices | phase1-simplefold3b-benchmark-slices | 3/3 provisional targets scored |
| Phase 1A frozen protocol | phase1a-protocol | 8-target computational freeze |
| Phase 1A manifest | CSV / JSON | published |
| Phase 1A prospective runs | phase1a-simplefold3b | 4/4 prospective targets scored |
| Phase 1A summary | CSV / JSON | generated from cards |
| Phase 1A selective repeats | repeat controls / JSON | complete for 11BC:C and 9K9Y:A |
| Second committee response | second-committee-response page | title, abstract, claim table, target rationale, and scorer detail hardened |
| lDDT scorer smoke test | pilot/scripts/test_lddt_scorer.py | passing in pilot/m3max-server/.venv |
| SimpleFold-3B model/runtime card | public card / environment JSON | Phase 1A route documented |
| Phase 1A biological review packet | public page / JSON | structured pre-review complete |
| Solo-authorship policy | solo-authorship page | Raúl solo author/corresponding author by default |
| Reviewer packet | reviewer-packet page | claim-to-evidence map drafted |
| Submission statements | submission-statements page | drafted |
| Public source release audit | public-release-audit page | public release complete |
| Public source release | v0.1.2-phase1a release | clean snapshot public |
| Archive DOI | 10.5281/zenodo.20275055 | source/data/PDF archived |
| Results discussion | manuscript discussion | updated from Phase 1A outcomes |
| Difficulty log | public difficulty log | active |
Remaining Before Stronger Submission Targets
| Gate | Blocking evidence | Expected output |
|---|---|---|
| External biological review | Optional domain-reviewer note on the structured Phase 1A packet before biological interpretation; acknowledge reviewers only with permission unless authorship criteria are met | signed reviewer note, acknowledgement permission, or explicit limitation |
| Phase 1B install cards | Commit, weights, license note, smoke command, and failure evidence for any added model beyond SimpleFold-3B | pilot/model-cards/*.md |
| Additional repeated-run expansion | Optional only; current first-paper repeat controls cover 13CZ:A, 11BC:C, and 9K9Y:A | only needed if reviewers demand broader repeat coverage |
| Venue figures and tables | Resize and caption generated figures/tables for the chosen venue template | submission PDF / preprint packet |
| Final discussion | Tighten interpretation after biological target review | updated manuscript |
| Companion autonomous Reader | Needed only if a later methods paper claims autonomous paper reading rather than manual protocol curation | reader_raw.txt, recipe.json, audit entry |
Submission Boundary
The final paper can claim measured SimpleFold-3B local-inference viability under this exact frozen protocol. It cannot claim Apple Silicon beats CUDA, cannot claim a new protein model, cannot claim official AlphaFold 3 evaluation, and cannot generalize to other model families beyond attempted evidence.