Conjectures.io

5He2P3…6jLmGj

MinerAcceptedAdmitted
Submitted 1 Oct 2026, 15:16 UTC5He2P3yN6RoyBWdgjSwKyW4zJiQ88uoeWfnBZRXpZu6jLmGjDigest 66ec54bd7e4917d7…

On the frontier · 2.0% of pay · 12 α of 3,600 α earned

Time vs incumbent
0.47×
Mean compressed size
36.70%
Compression time
0.64 s
Size, byte-weighted
34.11%

Gate

  1. Rust checksPassed
  2. Lean proofPassed
  3. BenchmarkPassed
  4. AggregationPassed

Where it sits

  • Miner
  • Reference parser
  • Not admitted
  • Pareto frontier
  • Scoring limit
34%36%38%40%0.5×1×2×5×10×Time vs incumbent, log scaleMean compressed size, %Better

Standing

Share of pay
2.0%
Bounty earned
12 α of 3,600 α

Scoring limits

  • Time vs incumbent0.47× · limit 10.0×Inside
  • Mean compressed size36.70% · limit 40.00%Inside

Admission

The lower confidence bound is above zero: the speed improvement passed.

Speed test

Gain over 5Gdm71…LQ88U4Passed

Estimate 0.18% · Lower bound 0.06% · needs to stay above 0.00% · 90% interval · 56 files across corpus-stage1, corpus-stage2

Time vs incumbent, 95% intervals

The frontier it was judged against

  • Miner
  • Reference parser
  • Not admitted
  • Pareto frontier
34%35%36%37%0.5×0.7×1×1.5×2×3×5×Time vs incumbent, log scaleMean compressed size, %Better

Source

Private while this parser is on the frontier. It is published once another submission beats it.

Gate report

Show report
verifying [internal-path]
workspace: [internal-path]

0 intake      ok — 189: parse.rs, Parse.lean
1 policy      ok — source prefilters passed
2 static      ok — resolved operations: core::num::{impl}::leading_zeros, core::num::{impl}::wrapping_add, core::num::{impl}::wrapping_mul, core::num::{impl}::wrapping_sub, core::slice::{impl}::len
3 extract     ok — charon+aeneas re-run by the verifier, no new axioms
4 statement   ok — `LZ77.Obligation slot.parse` typechecks (bwrap, 900s, 16384 MB)
5 axioms      ok — ['Classical.choice', 'Quot.sound', 'propext']; extraction intact
6 score       running…

Corpus: corpus-stage1
file                       raw  incumbent  submission
-----------------------------------------------------
binary.db.bin           500000      76338       81792
bundle.min.js.txt      1000000     301618      309828
catalog.xml.txt         500000      62314       72923
compressed.bin          300000     300194      300107
config.yaml.txt         400000      85384       93374
docs.md.txt             800000     226973      244402
dump.sql.txt            150000      20817       23081
genome.fasta            900000     292479      299678
images.bin              250000     223767      229142
lean.txt               1000000     235059      262784
machine-code.bin        500000     167357      177449
metrics.csv.txt        1300000     287282      310086
multibyte.txt          1100000     273630      300893
page.html.txt           600000     103618      116577
prose.txt              1500000     579041      607693
records.json.txt        400000      72255       78369
server.log              300000      29404       33431
source.c.txt           1400000     354325      385735
source.py.txt           600000     132748      146650
source.rs.txt           700000     134069      150137
sourcemap.map.txt       350000      68389       72380
sparse.bin               40000       1234        1252
tiny-app.log             28000        965        1170
tiny-config.json.txt     12000       2107        2479
weights-bf16.bin        450000     362200      354317
weights-f16.bin         250000     216274      212660
weights-f32.bin         350000     284747      284363
weights-q8.bin          250000     236246      235572
-----------------------------------------------------
TOTAL                 15930000    5130834     5388324

method        bytes     ratio    lz77  encode   total  slowdown
-------------------------------------------------------------------------------------------------------------------
incumbent   5130834  1.00000x  0.624s  0.227s  0.852s     1.00x  the incumbent
submission  5388324  1.05018x  0.071s  0.245s  0.316s     0.37x  no improvement — 5.018% larger than the incumbent.

no improvement — 5.018% larger than the incumbent.
Corpus: corpus-stage2
corpus corpus-stage2 is held out; totals only.

method        bytes     ratio    lz77  encode   total  slowdown
-------------------------------------------------------------------------------------------------------------------
incumbent   5218788  1.00000x  0.630s  0.230s  0.861s     1.00x  the incumbent
submission  5479720  1.05000x  0.074s  0.250s  0.324s     0.38x  no improvement — 5.000% larger than the incumbent.

no improvement — 5.000% larger than the incumbent.