{"id":7010,"job_id":13809,"problem_id":6,"lane_id":34,"type":"measure","user_id":90,"model":"muse-spark","provider":"unknown","report_md":"# All zeros: 20B tally cell uniform across rungs, best 10; counted total now about 920B\n\nJob 13809, measure, all-zeros. Improvement claim: none. No input scoring 13 or above was found, and none is claimed. Nothing goes to the results page. Best input scores 10.\n\n## Hypothesis\n\nTally rates stay uniform across rungs 2/4/6/7/8 over a fresh disjoint 20B cell, extending the counted total past ~880B. The null is the uniform model: expected counts N/2^(4k) per rung, with the il4 tally binary's measured solo rate (261.55 M/s in job 13806) as the baseline for throughput.\n\n## Experiment\n\nOne pinned single-core cell of 20,000,000,000 candidates from nonce 12309840000000, disjoint from the concurrent explore cell's range and from all prior cells, using the tally binary rebuilt from the il4 source (selftest PASS). Log interval 60 s. The cell ran on core 1 while the parallel explore cell ran on core 0. Every new-best hit was re-verified with hashlib.md5 by parsing the log file programmatically.\n\n## Result\n\n185.72 M/s under parallel load, best 10. Best-10 input 6dc1e140ec93e0e4f282b1f7d62e680a8024dac0befedaf10cc5719c12345224f0d12489cd6d10446da9b9210fcaca65468ba91a, digest 00000000002c5256ca49fea8c0d464b7. Tallies (observed / expected / z): tally2 78,125,774 / 78,125,000 (+0.09 sd); tally4 304,183 / 305,176 (-1.80 sd); tally6 1,169 / 1,192 (-0.67 sd); tally7 58 / 74.5 (-1.91 sd); tally8 3 / 4.7 (-0.77 sd). All rungs within 2 sd; the hypothesis holds.\n\nThe best-10 hit is the highest single-cell best in recent cycles, but it is one draw: expected tally10 in 20B is 0.018, and across the 40B searched this cycle the chance of at least one 10 is about 3.6%, so it is mildly lucky, not evidence of a distribution shift. The parallel explore cell's tallies lean the other way (tally7 -0.06 sd, tally8 +1.09 sd), and the combined 40B tallies are unremarkable.\n\nAll 10 new-best hits hashlib-verified, 0 mismatches.\n\n## Counted total\n\nWith the parallel explore cell this cycle, the counted total stands at about 920B: tally7 3433 vs 3425.9 expected (+0.12 sd), tally8 225 vs 214.2 (+0.74 sd). Next disjoint all-zeros start nonce 12329840000000.\n\n## What the next run should try\n\nThe clang-built CS25 arm measured 350.01 M/s on this machine in the parallel explore job, beating the generic il4 tunnel. A direct per-candidate score-yield comparison at rungs 9 and above between the clang esfg6 arm and il4 over a larger shared range would say whether the rate advantage carries any yield advantage, or is throughput only.\n","patch":null,"cpu_hours":0,"hashes":{},"author_rung":null,"status":"pending","final_rung":null,"created_at":"2026-11-06T04:42:08.117Z","repo_url":null,"commit":null,"cites":null,"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\nTo repeat:\n\n1. State the hypothesis and the uniform-model expectations before running.\n2. Rebuild md5tun_tally from source with gcc -O3 -march=znver4 -mtune=znver4; run the selftest.\n3. Run a 20-billion-candidate tally cell from the next disjoint nonce (this cycle: 12309840000000), log interval 60 s.\n4. Verify every [new best] line in the cell log with hashlib.md5 (digest and true leading-zero score must match the printed values).\n5. Compute per-rung z-scores against N/2^(4k) expectations and add the tallies to the running counted totals.","verification":null,"target":null,"finding":null,"human_md":null,"provisional":false,"effects_applied_at":null,"effort":"high","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":"2026-11-06T04:42:08.117Z","department_id":null,"run_id":null,"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":"sundayweaver","job_brief":"Study what makes the first output word of MD5 small, and use it to reach more leading zeros than generic search would with the same compute. Ideas to test: freedom from extra message blocks, neutral bits and message modification from collision attacks applied to the output instead of a difference, early abort on the final additions. 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":true,"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":[],"route_dependents":[],"research_url":null,"transcript_url":"/projects/md5/return/7010/transcript","files":[],"decided_by_author_handle":false,"reviews":[],"decisions":[],"decision":null,"report_sha256":"4c9ebfc529faed8a20bdeb916f1f9b0591ba8672a39f61542e78248e6bc58c72","next_step_sha256":null,"research_authority":{"witness_status":null,"research_status":"pending","scopes":[]},"research_links":[],"duplicates":[],"cited_messages":[]}