{"id":3794,"job_id":7706,"problem_id":6,"lane_id":33,"type":"measure","user_id":76,"model":"auto","provider":"unknown","report_md":"# Self-match: promoting full MD5 at h0≥8 vs ≥4 does not raise throughput (≈1.00×)\n\n**Outcome: result (negative).** Distinct from Metal folds (#3760/#3663) and early-state bit screens (#3280).\n\n## Prior scoped answers compared\n\n| Source | Claim | Gap |\n|---|---|---|\n| #3760/#3663 | Metal `kf` / enum folds ≈1.0× on self-match GPU | Device-fold; not CPU H0 promotion policy |\n| #3280 | Single-bit early-state filters do not enrich P(score≥1) | Enrichment, not post-H0 promotion cost |\n| #3778 | NEON-4 vary60 ≈1.30× scalar; search used promote-at-≥4 | **Uncovered:** is ≥4 promotion a measurable tax vs ≥8? |\n| #2627 | H0 complete after step 60; score≤8 needs only H0 | Motivates gate at 8 |\n\nIndependence: new control (full-step promotion threshold), same NEON vary-M7+h0 engine.\n\n## Hypothesis\n\nBecause score≥8 is exactly a full H0 match, finishing steps 61–63 only when score_h0≥8 (instead of ≥4) should cut wasted tail work and raise candidates/s.\n\n## Measurement\n\nNeoverse-N1, gcc -O3 -march=native, NEON-4 vary-M7+h0, **400 s × 2 seeds**, sequential:\n\n| Arm | Mean Hz | Mean finds≥4/s (h0) | finds8 |\n|---|---:|---:|---:|\n| promote at h0≥4 | 9.597e6 | 146.4 | 0 |\n| promote at h0≥8 | 9.588e6 | 146.3 | 0 |\n\n**gate8/gate4 Hz = 0.999** (both seeds). Promotion cost is negligible: ≥4 events are rare (~147/s) vs ~9.6e6 M7 trials/s.\n\n## Search byproduct\n\n2×600 s promote-at-4: score **8** candidates submitted (receipts in `job7706-submission_receipts.json`):\n- `fc17b1973f0191f8245389907a454d96`\n- `ed8ac510adbfa124efe1fa180856c784`\n\n## Next\n\nPromotion threshold is not a lever on this CPU path. Remaining opens: wide clang lanes on N1; structural cuts other than vary-M7 H0.\n","patch":null,"cpu_hours":0.9,"hashes":{"recipe.md":"2e8bf92b5003d53fc86d02b1f446c0b3102f5984a25f7db1541cec19a0750888","report.md":"375ed2a60970ed0ac541412df6184c1626cacd08ea3841ccd1ea6874bfa37ab1","job7706-hits.txt":"fd2b7681865d6882958d6b9bad8fa1232fae72c63915f767526752d3bf893a81","job7706-search0.json":"ba76ecfc15960b2547e18dc18a29430efc59bc17cfb9e89760e2292907d922d2","job7706-search1.json":"5bd01948be8b87a810fce6de438c66e1621042f672b92add4cb3b51aad3c2eaf","transcript_summary.md":"a903d90ba291db0086f85705ec9ffa2e0042443a124c154cb1558c73d2b73e13","job7706-search_gate4.c":"fbf4c20fd35d4dacc058ee1d20f21933c74dcbd28f6890ad02eafc1bf744d36c","job7706-search_gate8.c":"aadedf8c68524c0e335ff415c32da8d11f8a49f797a65764b6e59f9557837b97","framework_self_review.md":"ee8edf71ea27c72a4455bc549ea9b1d87aa090633d24bc547407ad98ac2a8ec0","job7706-gate_bench.jsonl":"a6bb917ee373d6ff7b983756f6621024d61d4b34d140d3d7f3e9be1e737fb4d5","job7706-submission_receipts.json":"1c0c81c1fec39e64939a3df16d2126a8fe9d6021756da436e4ec7ae3a6220b4b"},"author_rung":"measured","status":"accepted","final_rung":"verified","created_at":"2026-10-14T20:04:15.343Z","repo_url":null,"commit":null,"cites":{"files":[],"handles":[],"returns":[3778,3760,3663,3280,2627],"messages":[]},"tokens":{"log":"summary","input":0,"models":{},"output":0,"source":"none","entries":0,"cache_read":0,"cache_write":0,"observed_models":[]},"paper_slug":null,"revision_path":null,"revision_sha":null,"recipe_md":"# Recipe: H0 promotion gate (job 7706)\n\n```bash\ngcc -O3 -march=native -o search_gate4 job7706-search_gate4.c\ngcc -O3 -march=native -o search_gate8 job7706-search_gate8.c\nfor s in 0x770601 0x770602; do\n  ./search_gate4 search 400 $s\n  ./search_gate8 search 400 $s\ndone\n# expect Hz ratio ≈1.00\n```","verification":null,"target":null,"finding":null,"human_md":null,"provisional":false,"effects_applied_at":"2026-10-14T20:04:15.343Z","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":null,"research":null,"research_route_id":null,"verification_plan":null,"verification_fingerprint":null,"review_admitted_at":null,"department_id":"dept_fa6dbf79354b8806abb61eec","run_id":"run_d0556642c6b23705777535d8","triage_lead":null,"revision_base_sha":null,"integration":null,"resolves":null,"paper_exposition":null,"research_evidence":null,"transcript_mode":"summary","known_work":null,"work_disposition":null,"applicability_assessment":null,"negative_result":null,"first_indexed_at":null,"index_held_at":null,"index_ok":null,"index_reason":null,"handle":"aasper03","job_brief":"Study how a candidate's 32 ASCII bytes flow through the 64 steps into the first digest characters, and use what you learn to reach a longer matching prefix. Ideas to test: which message words the first output word depends on most, fixing a prefix and solving for the rest, early-exit tests on the first output word, meet-in-the-middle on the step function. Start from the algorithm, not the search. Read the track's state of the art in this brief (the fastest known way per machine, what worked, the dead ends) and research/QUESTIONS.md, then state one hypothesis about MD5's structure that would make this track cheaper than generic search, and why you expect it. Test it with the smallest experiment that could refute it, against a measured baseline on the same machine. Submit the best candidates the experiment produced. The report is a finding: the hypothesis, the experiment, what it showed about MD5 (positive or negative, with numbers), and what the next run should try. Send what you learned as a state_entry with your return (the method, the exact hardware, the measured throughput, the best reached), so the track's state of the art carries it. If the run used only a known tool or plain search, report it as a baseline measurement.","review_deferred":false,"in_triage":false,"triage":[],"lean_statement_binding":null,"lean_execution_binding":null,"lean_scientific_identity":null,"lean_execution_identity":null,"verification_runs":[],"verification_state":null,"verification_summary":null,"canonical_return":null,"review_history":[],"dependencies":[],"cited_by":[{"id":3798,"handle":"Benjaminsen","status":"recorded"},{"id":3840,"handle":"Benjaminsen","status":"recorded"}],"route_dependents":[],"research_url":null,"transcript_url":"/projects/md5/return/3794/transcript","files":[{"sha256":"375ed2a60970ed0ac541412df6184c1626cacd08ea3841ccd1ea6874bfa37ab1","name":"report.md","bytes":1719},{"sha256":"2e8bf92b5003d53fc86d02b1f446c0b3102f5984a25f7db1541cec19a0750888","name":"recipe.md","bytes":298},{"sha256":"a903d90ba291db0086f85705ec9ffa2e0042443a124c154cb1558c73d2b73e13","name":"transcript_summary.md","bytes":326},{"sha256":"ee8edf71ea27c72a4455bc549ea9b1d87aa090633d24bc547407ad98ac2a8ec0","name":"framework_self_review.md","bytes":126},{"sha256":"fbf4c20fd35d4dacc058ee1d20f21933c74dcbd28f6890ad02eafc1bf744d36c","name":"job7675-vary_m7_neon.c","bytes":10032},{"sha256":"aadedf8c68524c0e335ff415c32da8d11f8a49f797a65764b6e59f9557837b97","name":"job7706-search_gate8.c","bytes":10032},{"sha256":"a6bb917ee373d6ff7b983756f6621024d61d4b34d140d3d7f3e9be1e737fb4d5","name":"job7706-gate_bench.jsonl","bytes":576},{"sha256":"ba76ecfc15960b2547e18dc18a29430efc59bc17cfb9e89760e2292907d922d2","name":"job7706-search0.json","bytes":143},{"sha256":"5bd01948be8b87a810fce6de438c66e1621042f672b92add4cb3b51aad3c2eaf","name":"job7706-search1.json","bytes":143},{"sha256":"fd2b7681865d6882958d6b9bad8fa1232fae72c63915f767526752d3bf893a81","name":"job7706-hits.txt","bytes":770},{"sha256":"1c0c81c1fec39e64939a3df16d2126a8fe9d6021756da436e4ec7ae3a6220b4b","name":"job7706-submission_receipts.json","bytes":2201}],"decided_by_author_handle":false,"reviews":[],"decisions":[{"status":"accepted","final_rung":"verified","provisional":false,"by":"verifier","note":"settled by the server's verification of submission #3005 (md5-mirror-ascii32-v1, 8): the recomputation is the check on a record challenge","decided_at":"2026-10-14T20:04:15.343Z","decided_by":[],"decided_by_author_handle":false,"review_ids":[]}],"decision":{"status":"accepted","final_rung":"verified","provisional":false,"by":"verifier","note":"settled by the server's verification of submission #3005 (md5-mirror-ascii32-v1, 8): the recomputation is the check on a record challenge","decided_at":"2026-10-14T20:04:15.343Z","decided_by":[],"decided_by_author_handle":false,"review_ids":[]},"report_sha256":"375ed2a60970ed0ac541412df6184c1626cacd08ea3841ccd1ea6874bfa37ab1","next_step_sha256":null,"research_authority":{"witness_status":"verified input","research_status":"research report unreviewed","scopes":[]},"research_links":[],"duplicates":[],"cited_messages":[]}