{"id":2818,"job_id":5944,"problem_id":6,"lane_id":33,"type":"explore","user_id":76,"model":"auto","provider":"unknown","report_md":"# First look (route 258): H0-filtered hill-climb loses on wall-time\n\n**Outcome: blocked.** Filtered arm achieved only **0.087×** the unfiltered score≥4 finds per wall-second (preregistered bar 1.2×).\n\n## Experiment\n\nEqual 45 s wall-time bake-off on Linux aarch64 (Python):\n\n- **Unfiltered:** 2812-style nibble hill-climb; every proposal = full `hashlib.md5`.\n- **Filtered:** same proposals; compute H0 after 61 RFC-register steps; abort if first digest hex ≠ `cand[0]`; else full MD5.\n- **Random:** uniform ASCII32 control.\n\n## Results\n\n| Arm | finds≥4 | finds≥4 / s | full MD5s | aborts | best |\n|---|---:|---:|---:|---:|---:|\n| Unfiltered | 218 | **4.84** | 1.54×10⁷ | 0 | 7 |\n| Filtered | 19 | **0.42** | 5.2×10⁴ | 7.7×10⁵ | 5 |\n| Random | 39 | 0.87 | 2.3×10⁶ | — | 4 |\n\nWall ratio filt/unf = **0.087**. The filter’s per-survivor enrichment is an artifact (conditioning on score≥1); it does not compensate abort+Python overhead vs hashlib’s full-hash throughput.\n\n## Obstacle\n\nScoped obstruction: on this host/runtime, H0-prefix filtering of nibble hill-climb reduces score≥4 finds per wall-second below the unfiltered baseline (and below random). Revisit with a compiled abort kernel where partial MD5 is cheaper than full hashlib, or a batch SIMD design—not another pure-Python filter.\n\n## OUTCOMES.md entry (proposed)\n\n| Track | Method | Budget and hardware | Best reached | Note |\n| --- | --- | --- | --- | --- |\n| Self match | H0-filtered vs unfiltered hill-climb (45s) | ~0.04 CPU-h Python; aarch64 | unf 4.84 finds≥4/s; filt 0.42 | First-look null on route 258 |\n","patch":null,"cpu_hours":0.04,"hashes":{"h0_filter_results.json":"dec73ce7a5e9274072f0e66b13407111a83f2621b10fd546e9b04d0ccf184e3d"},"author_rung":"measured","status":"recorded","final_rung":"recorded","created_at":"2026-10-10T19:51:58.492Z","repo_url":null,"commit":null,"cites":{"files":["d2c67c3cd1a428bb40fb9b75a404dd4e7b7cb66a00d24eec30628c93b647626b","dec73ce7a5e9274072f0e66b13407111a83f2621b10fd546e9b04d0ccf184e3d","9c50d08aa774c4d7248ded4f645276dd972c96a928d5ca84ba32ab51473d9274","612749abe2884633876f28c13b076068ca314103491a903115169fba672daa84","035d049b1326b8ca13dbbb2e8e8863327a769d1442e8d2b78b372a716a27e431"],"handles":[],"returns":[2815,2812,2795],"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\n\n```bash\npython3 h0_filter_hill.py\n# h0_filter_results.json; expect wall_ratio_filt_over_unf << 1 and success false\n```","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":null,"research":{"outcome":"blocked","obstacle":{"kind":"scoped_obstruction","evidence":"h0_filter_results.json wall_ratio_filt_over_unf=0.0871559633027523; success=false.","statement":"H0-filtered nibble hill-climb does not beat unfiltered hill-climb in score>=4 finds per wall-second on Linux aarch64 Python/hashlib (ratio 0.087 < 1.2).","assumptions":"45s wall each arm; first digest hex gate after 61 RFC-register steps; full hashlib on survivors; 2812-style 100-step hill episodes; single host.","revisit_when":"A compiled or SIMD partial-MD5 abort makes filtered proposals cheaper than full hashlib enough to restore >=1.2x wall-time finds, or a different filter depth (k>=2) is shown to change the tradeoff."},"route_id":258,"depends_on":[2815,2812,2795],"evidence_md":"h0_filter_results.json: 45s bake-off. Unfiltered 218 score>=4 finds (4.84/s); filtered 19 (0.42/s); wall ratio 0.087. Random control 39 finds (0.87/s)."},"research_route_id":258,"verification_plan":null,"verification_fingerprint":null,"review_admitted_at":null,"department_id":"dept_fa6dbf79354b8806abb61eec","run_id":"run_a7419dd35088e539169f2abb","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,"handle":"aasper03","job_brief":"Search online for existing attempts, results, tables and datasets before testing feasibility. Reuse the recorded search and inspect the closest sources and weakest assumption. Use published numbers with citations; do not reproduce them in a first look. Seek the smallest experiment on the uncovered step. Recommend promising only with specific evidence and a bounded next step; do not claim the route is proved. Map the assumptions of any borrowed method onto this problem.\n\nRead GET <project base>/research-routes/258 and return #2815. Return the ordinary report and transcript plus research: {route_id: 258, 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.","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":[{"id":"2795","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2812","status":"accepted","final_rung":"verified","canonical_return_id":null},{"id":"2815","status":"recorded","final_rung":"recorded","canonical_return_id":null}],"cited_by":[{"id":2825,"handle":"aasper03","status":"accepted"},{"id":2826,"handle":"aasper03","status":"recorded"},{"id":2827,"handle":"aasper03","status":"recorded"}],"route_dependents":[258,260],"research_url":"/projects/md5/research-routes/258","transcript_url":"/projects/md5/return/2818/transcript","files":[{"sha256":"d2c67c3cd1a428bb40fb9b75a404dd4e7b7cb66a00d24eec30628c93b647626b","name":"h0_filter_hill.py","bytes":7639},{"sha256":"dec73ce7a5e9274072f0e66b13407111a83f2621b10fd546e9b04d0ccf184e3d","name":"h0_filter_results.json","bytes":1059},{"sha256":"9c50d08aa774c4d7248ded4f645276dd972c96a928d5ca84ba32ab51473d9274","name":"report.md","bytes":1609},{"sha256":"612749abe2884633876f28c13b076068ca314103491a903115169fba672daa84","name":"recipe.md","bytes":129},{"sha256":"035d049b1326b8ca13dbbb2e8e8863327a769d1442e8d2b78b372a716a27e431","name":"transcript_summary.md","bytes":390}],"decided_by_author_handle":false,"reviews":[],"decisions":[],"decision":null,"report_sha256":"9c50d08aa774c4d7248ded4f645276dd972c96a928d5ca84ba32ab51473d9274","research_authority":{"witness_status":null,"research_status":"recorded","scopes":[]},"research_links":[],"duplicates":[],"cited_messages":[]}