Science: prove2me
Status
Trial 8 is running, scaffold v9, since 16 Nov.
- nodes online (whole network)
- GPU slots busy now
- proofs accepted by prove2me, of submitted
- 7 of 7
- Qwen3.8 27B, an open-weight model, works on every task on the AIs@home network’s GPUs.
- A proof is submitted only after our own full Lean check passes it; prove2me’s check then decides.
- Each task is one attempt: an outline of a leaf, one of its holes, or putting the pieces together.
- Have an NVIDIA GPU with 16 GB or more? Run a node. It earns credit in BOINC’s unit, not money.
As of the site’s last build, 17 Nov, 20:00 UTC. The queue and the network show when the page can reach the gateway.
prove2me collects theorems from mathematical research papers, stated in Lean 4 and split into small pieces called leaves. Anyone, a person or an agent, may submit a proof of an open leaf, and prove2me’s own check decides: the Lean kernel and the axioms the proof uses.
Qwen3.8 27B, an open-weight model on the AIs@home network, works on leaves as an agent, with Lean and Mathlib at prove2me’s versions. Our own full check judges every attempt: the proof must build, state exactly the leaf’s theorem, and use no sorry and no axiom beyond Lean’s standard three. Only proofs that pass it are submitted, and since 12 November 2026 every one of them that proves an open leaf is. AIs@home is not part of prove2me.
Updated on
Results so far
- proofs accepted by prove2me, of 7 submitted
- 7
- open leaves proved, of 21 tried
- 5
- open leaves closed by reuse of a proof already on prove2me
- 2
- runs proved, of 48 in trials 1 to 7
- 15
Small numbers. The 5 open leaves proved, p2m-7f145a (O04), p2m-a146ce (F01), p2m-dbd69f (N03), p2m-aec6a2 (N10) and p2m-d00cfc (N07), show that the model can do it, not how often: p2m-7f145a (O04) (trial 3) was a short proof of a leaf nobody had tried yet, and p2m-a146ce (F01) (trial 5) took a new scaffold, two outlines and 19 tasks after trials 3 and 4 had failed on it. p2m-b1d59e (N02) and p2m-f181df (N05), closed by reuse, are not among them. On calibration leaves, which others had proved before, 10 of 15 runs succeeded. All runs together took 91.3 hours of model time, and the outline probes that chose leaves 4.0 more.
prove2me on the network
- requests today (UTC)
- of all the network’s requests today
- requests in all
- tokens computed in all
- GPU hours in all
Live from the gateway’s public statistics, updated UTC. The network counts prove2me on its own since 11 November 2026 (trial 4 on); the requests of trials 1 to 3 count under Tau Ceti’s project.
Submitted to prove2me
| Leaf | Theorem | From | Submitted | prove2me’s verdict |
|---|---|---|---|---|
| p2m-7f145aO04 | SeqPostedPricing.MatroidSPM.accept_prob_eq_quantile§4: offered the price F_i^{-1}(1-q_i), agent i accepts with probability exactly q_i | Trial 3 | Accepted06:51 UTC | |
| p2m-a146ceF01 | CodedSorting.proposition_7_1_transposeProposition 7.1 (rearrangement part; companion: proposition_7_1_correctness) — \pi and \pi^{-1} are three adjacent field interchanges | Trial 5 | Accepted20:10 UTC | |
| p2m-b1d59eN02 | occupation_bind_sections_integrable_localsame statement as AvramDividend.Classical.avram_chat24_occupation_bind_sections_integrable_public by WillR | Closed by reusefound while preparing trial 6 | Accepted21:55 UTC | |
| p2m-f181dfN05 | NestedSeatAlloc.IntPolicy.expRevenue_one_right_derivativeEquation (27) right derivative for one-class expected revenuepart of NestedSeatAlloc.IntPolicy.eq27_er1_clbi by mikedeng1 | Closed by reusefound during trial 6 | Accepted10:51 UTC | |
| p2m-dbd69fN03 | PerturbedWF.eq_2_3Eq. (2.3): \langle v,x-y\rangle\le\|x-y\| for v\in B^* and x,y\in\mathcal C | Trial 6 | Accepted14:51 UTC | |
| p2m-aec6a2N10 | thm_5_6Chapter 5, thm_5_6 | Trial 7 | Accepted15:09 UTC | |
| p2m-d00cfcN07 | MatchIndiff.PublicSM.lemma_2_1Lemma 2.1 — accepted profiles of a fixed board form a product, and a row and board fix the decoded component | Trial 7 | Accepted20:49 UTC |
p2m-7f145a (O04) had been open for 13 days with no submission from anyone. Its proof is short: the distribution has a density, so a single price carries no probability mass. p2m-a146ce (F01) was still open when we submitted it. Its proof follows the leaf’s own statement: the index is split into four fields, each of the outline’s six steps was proved by a task of its own, and the pieces were put together into one proof, which our full check and then prove2me accepted. prove2me’s theorem pages show both leaves as proved. Calibration leaves are never submitted: they were proved before.
p2m-b1d59e (N02) was closed by reuse. The leaf is the same statement as a theorem that WillR had already proved on prove2me under another name. Our index of prove2me’s library found the match while we prepared trial 6, and the model’s proof is one line that imports that theorem. Our full check and then prove2me accepted it, but none of the mathematics is ours, so it is not counted among the open leaves proved.
p2m-f181df (N05) was closed by reuse. The leaf is one part of a theorem that mikedeng1 had already proved on prove2me: its conclusion contains the leaf’s statement. Our index of prove2me’s library offered that theorem as a similar statement during trial 6, and the model’s proof is one line that uses it. Our full check and then prove2me accepted it, but none of the mathematics is ours, so it is not counted among the open leaves proved.
Working on now
Trial 8, since 16 November 2026 · scaffold v9
Budget: one round of 110 steps / 110 min for the outline and for each hole, then one fix round of 20 / 20; up to 2 tries per task.
- p2m-3ad60fC07WilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Queued: outline queued
- p2m-754bbcN11RAREPricing.Existence.observation_3Queued: outline queued
- p2m-527634N12DistLTLSynth.Char.lemma_2_3Queued: outline queued
- p2m-03d4b4N13DynApproxIdx.Main.fact_7_6Queued: outline queued
- p2m-580c57N14DuffieHedge.Optimal.F_posQueued: outline queued
- p2m-fe7aa1N15MIBF.Balanced.full_theorem_6_4Queued: outline queued
Every leaf
All 35 leaves tried so far, one row per leaf over every trial that ran it. A leaf’s id here is p2m- and the start of prove2me’s own id for the theorem; the id the trials used (C11, F01, …) is shown under it. Its outcome is the best of its runs: accepted by prove2me, closed by reuse, proved, not proved, stopped by us, or not scored.
| Role | Leaves | Runs | Runs proved | Proofs accepted | Closed by reuse | Model hours | Tokens computed |
|---|---|---|---|---|---|---|---|
| Calibration | 12 | 15 | 10 | 0 | 0 | 10.9 | 2.3M |
| Open | 23 | 29 | 4 | 4 | 2 | 56.4 | 13.8M |
| Comparison | 3 | 4 | 1 | 1 | 0 | 24.0 | 4.9M |
| Outline probes | 11 | – | – | – | – | 4.0 | 809k |
| All | 35 | 48 | 15 | 5 | 2 | 95.3 | 21.7M |
A comparison leaf is an open leaf run again to compare a new scaffold with the trials before it. A leaf can have more than one role over its trials, so the leaves by role add up to more than 35. A proof accepted counts under the role of the run that proved it; a leaf closed by reuse has no run. An outline probe is one short outline task on a candidate leaf, run to choose a trial’s leaves: it is no run, but its model time and tokens count in the totals.
All 35 leaves
Showing 35 of 35
| Role | Trials | ||||||
|---|---|---|---|---|---|---|---|
| p2m-6c35d7A01 | WilliamsonShmoys.connected_planar_supergraph_from_step__p2m_f2188066_bdee_41e3_86df_42250492e63d_a4 | Open | 1, 2 | Not proved0 of 2 runs proved | 92.9 | 209k | |
| p2m-2499fdA02 | QAlgorithms.Papers.Independent.Intakethird100.P20.regularized_base_is_quasipoint | Open | 1, 2 | Not proved0 of 2 runs proved | 89.9 | 239k | |
| p2m-12ea27A03 | QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivial | Open, Comparison | 1, 2, 4 | Not proved0 of 3 runs proved | 155.0 | 755k | |
| p2m-f10c9dA04 | SongZipkinFluct.Monotone.corollary2 | Open, Comparison | 1, 2, 4 | Not proved0 of 3 runs proved | 176.4 | 511k | |
| p2m-35e2daB01 | FasterSFM.lemma_13 | Open | 1 | Not proved | 29.0 | 87k | |
| p2m-ca216eB02 | OAI.MUB6.Certificate.GramData.Block3.entry_11_11 | Open | 1 | Not scored | 0.2 | 54k | |
| p2m-87ebddB03 | TenChallenges.challenge_04_r3 | Open | 1 | Not proved | 16.1 | 52k | |
| p2m-5e7987B04 | FalconerPath.completion | Open | 1 | Not proved | 29.0 | 88k | |
| p2m-42603fB05 | QuantumMaxCutEnergy.exists_0562_approximation | Open | 1 | Not scored | 19.2 | 68k | |
| p2m-6ed316B06 | QMCLeafB.SnapLedger.ledger_round_ok | Open | 1 | Not proved | 9.9 | 68k | |
| p2m-7ab928C01 | DPDeconstructed.lemma_6_1 | Calibration | 2 | Proved | 20.8 | 57k | |
| p2m-81b93aC02 | MarkovProcess.measurable_finTail | Calibration | 2 | Proved | 11.6 | 19k | |
| p2m-12f452C03 | OnlineCombOpt.Exp2Full.theorem_1_full_domain | Calibration | 2 | Not proved | 65.1 | 176k | |
| p2m-12087dC04 | WilliamsonShmoys.face_accounting_implies_edge_bound | Calibration | 2 | Proved | 5.2 | 13k | |
| p2m-8b8abcC05 | Groundwork.Colouring.card_prescribed | Calibration | 2 | Not proved | 49.5 | 157k | |
| p2m-dfe8e4C06 | BianchiIX.measurableEmbedding_gaussWord | Calibration | 2 | Proved | 61.7 | 205k | |
| p2m-a146ceF01 | CodedSorting.proposition_7_1_transpose | Open, Comparison | 3, 4, 5 | Accepted by prove2me1 of 3 runs proved | 1369.2 | 4.6M | |
| p2m-c3c3c4F02 | APSCounting893.Main.generic_linear_noether_normalization | Open | 3 | Not proved | 78.0 | 516k | |
| p2m-8503e6F03 | BirthdayEntropy.Soundness.pair_avoidance | Open | 3 | Not proved | 69.8 | 209k | |
| p2m-c0eca8O01 | HolevoCQ.decoder_subPOVM | Open | 3, 4 | Not proved0 of 2 runs proved | 143.7 | 669k | |
| p2m-83aae8O02 | BBPSSW.Distill.typical_set_sqrt | Open | 3, 4, 5 | Not proved0 of 3 runs proved | 821.7 | 3.1M | |
| p2m-d1f097O03 | DigitalPrivacy.DataStrategy.proof_thm1_case_psi_lt_Prho | Open | 3, 4 | Not proved0 of 2 runs proved | 165.8 | 482k | |
| p2m-7f145aO04 | SeqPostedPricing.MatroidSPM.accept_prob_eq_quantile | Open | 3 | Accepted by prove2me | 53.9 | 132k | |
| p2m-3ad60fC07 | WilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4 | Calibration | 3, 6, 7 | Proved3 of 3 runs proved | 143.3 | 502k | |
| p2m-eb3303C08 | AvramDividend.Classical.avram_ubv_finite9_uniformIntegrable_one_of_common_integrable_dominator | Calibration | 3 | Not proved | 42.4 | 175k | |
| p2m-2b384dC09 | FiniteEnergyPlane.integral_norm_sq_le_of_skew | Calibration | 3 | Not proved | 57.5 | 242k | |
| p2m-479463C10 | BookProof.ChapterF3.dressed_vacuum_bessel | Calibration | 3 | Proved | 9.0 | 63k | |
| p2m-63eff5C11 | Ideal.Factors.quotientResidueEquiv_mk | Calibration | 4 | Proved | 3.7 | 21k | |
| p2m-6f7dceC12 | InsiderDonations.price_recast | Calibration | 4, 5 | Proved1 of 2 runs proved | 185.7 | 627k | |
| p2m-dbd69fN03 | PerturbedWF.eq_2_3 | Open | 6 | Accepted by prove2me | 165.6 | 695k | |
| p2m-aec6a2N10 | thm_5_6 | Open | 7 | Accepted by prove2me | 20.9 | 138k | |
| p2m-d00cfcN07 | MatchIndiff.PublicSM.lemma_2_1 | Open | 7 | Accepted by prove2me | 47.3 | 270k | |
| p2m-115f6cN08 | QAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1 | Open | 7 | Not proved | 1267.6 | 5.8M | |
| p2m-b1d59eN02 | occupation_bind_sections_integrable_local | Open | – | Closed by reuse | – | – | |
| p2m-f181dfN05 | NestedSeatAlloc.IntPolicy.expRevenue_one_right_derivative | Open | – | Closed by reuse | – | – |
Every task
Every task of every trial and of the outline probes, grouped by leaf and outline; a run of trials 1 to 4 is one task. From trial 6 on, the tasks come live from the gateway’s records, each trial as soon as its tasks are made.
All tasks
Showing 136 of 136
| Leaf | Task | Result | Work | Finished |
|---|---|---|---|---|
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 7 | Outline 1 | Written, 1 hole | 73 steps · 63 min · 157k | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 7 | Hole 1 (outline 1) | Proved | 24 steps · 8.6 min · 45k | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 7 | Assembly (outline 1) | Proved: our full check passed | – | – |
| p2m-aec6a2thm_5_6Trial 7 | Outline 1 | Written, 2 holes | 26 steps · 6.9 min · 68k | |
| p2m-aec6a2thm_5_6Trial 7 | Hole 1 (outline 1) | Proved | 22 steps · 8.0 min · 37k | |
| p2m-aec6a2thm_5_6Trial 7 | In place (outline 1) | Filled in, checked by the assembly | 24 steps · 6.0 min · 34k | |
| p2m-aec6a2thm_5_6Trial 7 | Assembly (outline 1) | Proved: our full check passed | – | – |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Outline 1 | Degenerate outline | 23 steps · 13 min · 76k | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Outline 2replaces outline 1 | Degenerate outline | 30 steps · 22 min · 103k | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Outline 3replaces outline 2 | Written, 2 holes | 17 steps · 9.2 min · 64k | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Hole 1 (outline 3) | Proved | 9 steps · 2.1 min · 15k | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Hole 2 (outline 3) | Proved | 7 steps · 1.2 min · 13k | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 | Assembly (outline 3) | Proved: our full check passed | – | – |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Outline 1 | Written, 8 holes | 36 steps · 19 min · 106k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 2 (outline 1) | Proved | 11 steps · 3.2 min · 14k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 3 (outline 1) | Proved | 20 steps · 5.0 min · 49k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 4 (outline 1) | Proved | 30 steps · 8.6 min · 47k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 5 (outline 1) | Proved | 31 steps · 13 min · 69k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 6 (outline 1) | A sorry left | 128 steps · 48 min · 234k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 6 (outline 1) · try 2retry of try 1 | Disproved: the hole is false | 97 steps · 60 min · 298k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 7 (outline 1) | Did not compile | 128 steps · 67 min · 301k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 7 (outline 1) · try 2retry of try 1 | A sorry left | 112 steps · 95 min · 527k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 8 (outline 1) | Disproved: the hole is false | 128 steps · 50 min · 258k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Outline 2replaces outline 1 | Written, 3 holes | 57 steps · 72 min · 241k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 1 (outline 2) | Proved | 38 steps · 28 min · 115k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 2 (outline 2) | Proved | 21 steps · 15 min · 50k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 3 (outline 2) | A sorry left | 106 steps · 116 min · 624k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 3 (outline 2) · try 2retry of try 1 | A sorry left | 80 steps · 129 min · 666k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Outline 3replaces outline 2 | Written, 9 holes | 99 steps · 66 min · 241k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 1 (outline 3) | Proved | 24 steps · 4.4 min · 52k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 2 (outline 3) | A sorry left | 40 steps · 130 min · 517k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 2 (outline 3) · try 2retry of try 1 | A sorry left | 119 steps · 110 min · 395k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 3 (outline 3) | Proved | 31 steps · 30 min · 125k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 4 (outline 3) | Proved | 66 steps · 27 min · 119k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 5 (outline 3) | Proved | 65 steps · 27 min · 90k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 6 (outline 3) | Proved | 104 steps · 44 min · 200k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 7 (outline 3) | Proved | 74 steps · 34 min · 135k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 8 (outline 3) | A sorry left | 19 steps · 60 min · 265k | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 8 (outline 3) · try 2retry of try 1 | Cancelled (outline refuted) | – | – |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 | Hole 9 (outline 3) | Disproved: the hole is false | 13 steps · 6.3 min · 39k | |
| p2m-51d621QiFixedDesign.Concentration.lemma_4Trial 7 probes | Probe: outline | A sorry left | 35 steps · 45 min | |
| p2m-d00cfcMatchIndiff.PublicSM.lemma_2_1Trial 7 probes | Probe: outline | Degenerate outline | 18 steps · 14 min | |
| p2m-115f6cQAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Trial 7 probes | Probe: outline | Cancelled | – | |
| p2m-17cf43MatchIndiff.PublicSM.lemma_D_9Trial 7 probes | Probe: outline | Cancelled | – | |
| p2m-aec6a2thm_5_6Trial 7 probes | Probe: outline | Cancelled | – | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 6 | Outline 1 | Written, 3 holes | 19 steps · 6.6 min · 55k | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 6 | Hole 2 (outline 1) | Proved | 6 steps · 1.7 min · 14k | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 6 | Assembly (outline 1) | Proved: our full check passed | – | – |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Outline 1 | Written, 2 holes | 59 steps · 27 min · 106k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Hole 2 (outline 1) | Our check failed | 19 steps · 7.0 min · 39k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Hole 2 (outline 1) · try 2retry of try 1 | Cancelled (outline replaced) | – | – |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | In place (outline 1) | Cancelled (outline replaced) | – | – |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Outline 2replaces outline 1 | Written, 4 holes | 28 steps · 13 min · 50k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Hole 1 (outline 2) | Proved | 21 steps · 4.8 min · 21k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Hole 3 (outline 2) | Proved | 47 steps · 44 min · 176k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Hole 4 (outline 2) | Proved | 15 steps · 7.6 min · 28k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | In place (outline 2) | Filled in, checked by the assembly | 118 steps · 62 min · 274k | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 | Assembly (outline 2) | Proved: our full check passed | – | – |
| p2m-63eff5Ideal.Factors.quotientResidueEquiv_mkTrial 6 probes | Probe: outline | Proved: our full check passed | 19 steps · 14 min | |
| p2m-26d409predictable_section_measurable_localTrial 6 probes | Probe: outline | Written, 2 holes | 41 steps · 39 min | |
| p2m-b1d59eoccupation_bind_sections_integrable_localTrial 6 probes | Probe: outline | Proved: our full check passed | 11 steps · 4.3 min | |
| p2m-dbd69fPerturbedWF.eq_2_3Trial 6 probes | Probe: outline | Did not compile | 33 steps · 45 min | |
| p2m-c1b761PochetWolsey.Variants.sales_extreme_pointsTrial 6 probes | Probe: outline | Did not compile | 16 steps · 45 min | |
| p2m-f181dfNestedSeatAlloc.IntPolicy.expRevenue_one_right_derivativeTrial 6 probes | Probe: outline | Outline written | 43 steps · 33 min | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Outline 1 | Written, 7 holes | 29 steps · 25 min · 135k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 1 (outline 1) | Proved | 25 steps · 20 min · 80k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 2 (outline 1) | Proved | 35 steps · 39 min · 235k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 3 (outline 1) | Proved | 32 steps · 20 min · 104k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 4 (outline 1) | Proved | 81 steps · 94 min · 303k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 5 (outline 1) | Proved | 20 steps · 6.3 min · 29k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 6 (outline 1) | Did not compile | 92 steps · 130 min · 364k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 6 (outline 1) · try 2retry of try 1 | Did not compile | 65 steps · 131 min · 363k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 7 (outline 1) | Did not compile | 45 steps · 130 min · 335k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 7 (outline 1) · try 2retry of try 1 | A sorry left | 105 steps · 130 min · 278k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Outline 2replaces outline 1 | Written, 6 holes | 27 steps · 33 min · 87k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 1 (outline 2) | Proved | 31 steps · 39 min · 102k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 2 (outline 2) | Proved | 36 steps · 56 min · 141k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 3 (outline 2) | Did not compile | 89 steps · 130 min · 428k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 3 (outline 2) · try 2retry of try 1 | Proved | 38 steps · 90 min · 245k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 4 (outline 2) | Proved | 38 steps · 67 min · 166k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 5 (outline 2) | Proved | 18 steps · 14 min · 38k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Hole 6 (outline 2) | Proved | 39 steps · 55 min · 218k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 5 | Assembly (outline 2) | Proved: our full check passed | – | – |
| p2m-6f7dceInsiderDonations.price_recastTrial 5 | Outline 1 | Written, 1 hole | 66 steps · 110 min · 371k | |
| p2m-6f7dceInsiderDonations.price_recastTrial 5 | Hole 1 (outline 1) | Proved | 25 steps · 25 min · 97k | |
| p2m-6f7dceInsiderDonations.price_recastTrial 5 | Assembly (outline 1) | Assembly did not check | – | – |
| p2m-6f7dceInsiderDonations.price_recastTrial 5 | Assembly (outline 1) · try 2retry of try 1 | Proved: our full check passed | – | – |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Outline 1 | Written, 6 holes | 32 steps · 56 min · 115k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 1 (outline 1) | Stopped by a fault | 18 steps · 19 min · 77k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 1 (outline 1) · try 2retry of try 1 | Stopped, no progress | 91 steps · 110 min · 300k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 2 (outline 1) | Stopped, no progress | 25 steps · 110 min · 260k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 2 (outline 1) · try 2retry of try 1 | A sorry left | 127 steps · 88 min · 268k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 3 (outline 1) | A sorry left | 82 steps · 130 min · 403k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | In place (outline 1) | Stopped, no progress | 100 steps · 68 min · 335k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Outline 2replaces outline 1 | Written, 12 holes | 44 steps · 61 min · 162k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | Hole 1 (outline 2) | Cancelled | – | – |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 5 | In place (outline 2) | Cancelled | – | – |
| p2m-63eff5Ideal.Factors.quotientResidueEquiv_mkTrial 4 | Run | Proved: our full check passed | 20 steps · 3.7 min · 21k | |
| p2m-6f7dceInsiderDonations.price_recastTrial 4 | Run | A sorry left | 15 steps · 50 min · 159k | |
| p2m-12ea27QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialTrial 4 | Run | Stopped, no progress | 13 steps · 65 min · 311k | |
| p2m-f10c9dSongZipkinFluct.Monotone.corollary2Trial 4 | Run | A sorry left | 43 steps · 83 min · 232k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 4 | Run | A sorry left | 45 steps · 81 min · 666k | |
| p2m-c0eca8HolevoCQ.decoder_subPOVMTrial 4 | Run | Stopped, no progress | 52 steps · 65 min · 348k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 4 | Run | A sorry left | 48 steps · 85 min · 268k | |
| p2m-d1f097DigitalPrivacy.DataStrategy.proof_thm1_case_psi_lt_PrhoTrial 4 | Run | A sorry left | 39 steps · 85 min · 249k | |
| p2m-a146ceCodedSorting.proposition_7_1_transposeTrial 3 | Run | A sorry left | 59 steps · 79 min · 270k | |
| p2m-c3c3c4APSCounting893.Main.generic_linear_noether_normalizationTrial 3 | Run | A sorry left | 42 steps · 78 min · 516k | |
| p2m-8503e6BirthdayEntropy.Soundness.pair_avoidanceTrial 3 | Run | A sorry left | 19 steps · 70 min · 209k | |
| p2m-c0eca8HolevoCQ.decoder_subPOVMTrial 3 | Run | Did not compile | 59 steps · 79 min · 321k | |
| p2m-83aae8BBPSSW.Distill.typical_set_sqrtTrial 3 | Run | A sorry left | 49 steps · 95 min · 874k | |
| p2m-d1f097DigitalPrivacy.DataStrategy.proof_thm1_case_psi_lt_PrhoTrial 3 | Run | Did not compile | 35 steps · 81 min · 233k | |
| p2m-7f145aSeqPostedPricing.MatroidSPM.accept_prob_eq_quantileTrial 3 | Run | Proved: our full check passed | 47 steps · 54 min · 132k | |
| p2m-3ad60fWilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4Trial 3 | Run | Proved: our full check passed | 60 steps · 64 min · 231k | |
| p2m-eb3303AvramDividend.Classical.avram_ubv_finite9_uniformIntegrable_one_of_common_integrable_dominatorTrial 3 | Run | Did not compile | 65 steps · 42 min · 175k | |
| p2m-2b384dFiniteEnergyPlane.integral_norm_sq_le_of_skewTrial 3 | Run | A sorry left | 48 steps · 58 min · 242k | |
| p2m-479463BookProof.ChapterF3.dressed_vacuum_besselTrial 3 | Run | Proved: our full check passed | 21 steps · 9.0 min · 63k | |
| p2m-7ab928DPDeconstructed.lemma_6_1Trial 2 | Run | Proved: our full check passed | 27 steps · 21 min · 57k | |
| p2m-81b93aMarkovProcess.measurable_finTailTrial 2 | Run | Proved: our full check passed | 17 steps · 12 min · 19k | |
| p2m-12f452OnlineCombOpt.Exp2Full.theorem_1_full_domainTrial 2 | Run | A sorry left | 27 steps · 65 min · 176k | |
| p2m-12087dWilliamsonShmoys.face_accounting_implies_edge_boundTrial 2 | Run | Proved: our full check passed | 12 steps · 5.2 min · 13k | |
| p2m-8b8abcGroundwork.Colouring.card_prescribedTrial 2 | Run | A sorry left | 68 steps · 50 min · 157k | |
| p2m-dfe8e4BianchiIX.measurableEmbedding_gaussWordTrial 2 | Run | Proved: our full check passed | 57 steps · 62 min · 205k | |
| p2m-6c35d7WilliamsonShmoys.connected_planar_supergraph_from_step__p2m_f2188066_bdee_41e3_86df_42250492e63d_a4Trial 2 | Run | Did not compile | 45 steps · 68 min · 143k | |
| p2m-2499fdQAlgorithms.Papers.Independent.Intakethird100.P20.regularized_base_is_quasipointTrial 2 | Run | Did not compile | 46 steps · 64 min · 162k | |
| p2m-12ea27QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialTrial 2 | Run | A sorry left | 36 steps · 60 min · 323k | |
| p2m-f10c9dSongZipkinFluct.Monotone.corollary2Trial 2 | Run | A sorry left | 45 steps · 63 min · 196k | |
| p2m-6c35d7WilliamsonShmoys.connected_planar_supergraph_from_step__p2m_f2188066_bdee_41e3_86df_42250492e63d_a4Trial 1 | Run | Did not compile | 29 steps · 25 min · 66k | |
| p2m-2499fdQAlgorithms.Papers.Independent.Intakethird100.P20.regularized_base_is_quasipointTrial 1 | Run | Did not compile | 7 steps · 26 min · 78k | |
| p2m-12ea27QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialTrial 1 | Run | No solution file | 25 steps · 30 min · 121k | |
| p2m-f10c9dSongZipkinFluct.Monotone.corollary2Trial 1 | Run | A sorry left | 20 steps · 30 min · 83k | |
| p2m-35e2daFasterSFM.lemma_13Trial 1 | Run | No solution file | 14 steps · 29 min · 87k | |
| p2m-ca216eOAI.MUB6.Certificate.GramData.Block3.entry_11_11Trial 1 | Run | No solution file | 3 steps · 0.2 min · 54k | |
| p2m-87ebddTenChallenges.challenge_04_r3Trial 1 | Run | No solution file | 29 steps · 16 min · 52k | |
| p2m-5e7987FalconerPath.completionTrial 1 | Run | No solution file | 26 steps · 29 min · 88k | |
| p2m-42603fQuantumMaxCutEnergy.exists_0562_approximationTrial 1 | Run | No solution file | 19 steps · 19 min · 68k | |
| p2m-6ed316QMCLeafB.SnapLedger.ledger_round_okTrial 1 | Run | No solution file | 30 steps · 9.9 min · 68k |
Trials
Each trial changed the agent’s scaffold (its instructions, tools and rounds) from what the one before taught, so the trials are steps, not repeats. Tap a trial for every leaf it ran.
| Trial | Dates | Scaffold | Proved | Calibration | Open | Model hours | Tokens computed |
|---|---|---|---|---|---|---|---|
| Trial 1 | v1 | 0 of 10 | – | 0 of 10 | 3.6 | 763k | |
| Trial 2 | v2 | 4 of 10 | 4 of 6 | 0 of 4 | 7.8 | 1.5M | |
| Trial 3 | v3 | 3 of 11 | 2 of 4 | 1 of 7 | 11.8 | 3.3M | |
| Trial 4 | v4 | 1 of 8 | 1 of 2 | 0 of 6 | 8.7 | 2.3M | |
| Trial 5 | v5 | 2 of 3 | 1 of 1 | 1 of 2 | 33.1 | 6.0M | |
| Trial 6 probes | v6 | – | – | – | 3.0 | 558k | |
| Trial 6 | v7 | 2 of 2 | 1 of 1 | 1 of 1 | 2.9 | 763k | |
| Trial 7 probes | v7 | – | – | – | 1.0 | 251k | |
| Trial 7 | v8 | 3 of 4 | 1 of 1 | 2 of 3 | 23.5 | 6.4M | |
| All | 15 of 48 | 10 of 15 | 5 of 33 | 95.3 | 21.7M |
Trial 1 · 7 November 2026 · 0 of 10 proved
No leaf proved. Seven of ten runs never wrote a solution file: the agent spent most of its steps searching library source. Two runs were cut off by the gateway's personal-data filter, which took long integers in the mathematics for card numbers; that filter is fixed and the two runs are not scored.
Budget per leaf: one round of 30 steps / 30 model minutes.
| Leaf | Theorem | Kind | Result | Rounds | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-6c35d7A01 | WilliamsonShmoys.connected_planar_supergraph_from_step__p2m_f2188066_bdee_41e3_86df_42250492e63d_a4Finite maximality turns strict planar extensions into connectivity | Open | Not proved: did not compile | 1 | 24.8 | 66k |
| p2m-2499fdA02 | QAlgorithms.Papers.Independent.Intakethird100.P20.regularized_base_is_quasipointLemma 3.4 | Open | Not proved: did not compile | 1 | 26.4 | 78k |
| p2m-12ea27A03 | QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialProposition 3.1 | Open | Not proved: no solution file | 1 | 29.5 | 121k |
| p2m-f10c9dA04 | SongZipkinFluct.Monotone.corollary2Corollary 2: the finite-horizon basestock levels y^*_n(i) converge as n\to\infty | Open | Not proved: a sorry left | 1 | 30.0 | 83k |
| p2m-35e2daB01 | FasterSFM.lemma_13Lemma 13 — inserted and cutting-plane row bounds: k\le k_n, Q_n<Kp_n^2, \chi_n<1 | Open | Not proved: no solution file | 1 | 29.0 | 87k |
| p2m-ca216eB02 | OAI.MUB6.Certificate.GramData.Block3.entry_11_11OAI.MUB6.Certificate.GramData.Block3.entry_11_11: intermediate lemma from OpenAI, The maximum number of mutually unbiased bases in dimension six | Open | Not scoredcut off by a false positive of the gateway's personal-data filter (since fixed) | 1 | 0.2 | 54k |
| p2m-87ebddB03 | TenChallenges.challenge_04_r3Sidorenko's inequality for sides of size at most 3 | Open | Not proved: no solution file | 1 | 16.1 | 52k |
| p2m-5e7987B04 | FalconerPath.completionA nonzero absolutely continuous measure on the original distance set | Open | Not proved: no solution file | 1 | 29.0 | 88k |
| p2m-42603fB05 | QuantumMaxCutEnergy.exists_0562_approximationClassical 0.562 approximation of Quantum Max-Cut energy | Open | Not scoredcut off by a false positive of the gateway's personal-data filter (since fixed) | 1 | 19.2 | 68k |
| p2m-6ed316B06 | QMCLeafB.SnapLedger.ledger_round_okSnap displacement versus the recorded margin (margin–ledger joint) | Open | Not proved: no solution file | 1 | 9.9 | 68k |
Trial 2 · 8 November 2026 · 4 of 10 proved
A new scaffold: write the solution file early, build after every edit, ask Lean instead of searching, and up to two fix rounds with Lean's errors. 4 of 6 calibration leaves proved, 3 of them in the first round; none of the 4 open leaves.
Budget per leaf: 30 steps / 30 min, then up to 2 fix rounds of 20 / 20.
| Leaf | Theorem | Kind | Result | Rounds | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-7ab928C01 | DPDeconstructed.lemma_6_1Lemma 6.1 — fixed points and iterates pass around the cycle \tau_0,\tau_1,\tau_2 | Calibration | Proved | 1 | 20.8 | 57k |
| p2m-81b93aC02 | MarkovProcess.measurable_finTailMeasurability of dropping the first finite coordinate | Calibration | Proved | 1 | 11.6 | 19k |
| p2m-12f452C03 | OnlineCombOpt.Exp2Full.theorem_1_full_domainTheorem 1 — full-domain EXP2 regret lower bound (corrected: excludes dimension one) | Calibration | Not proved: a sorry left | 3 | 65.1 | 176k |
| p2m-12087dC04 | WilliamsonShmoys.face_accounting_implies_edge_boundFace accounting implies the planar edge bound | Calibration | Proved | 1 | 5.2 | 13k |
| p2m-8b8abcC05 | Groundwork.Colouring.card_prescribedColouring: card prescribed | Calibration | Not proved: a sorry left | 3 | 49.5 | 157k |
| p2m-dfe8e4C06 | BianchiIX.measurableEmbedding_gaussWordMeasurable embeddings defined by finite Gauss words | Calibration | Proved in round 3 | 3 | 61.7 | 205k |
| p2m-6c35d7A01 | WilliamsonShmoys.connected_planar_supergraph_from_step__p2m_f2188066_bdee_41e3_86df_42250492e63d_a4Finite maximality turns strict planar extensions into connectivity | Opentried before in trial 1 | Not proved: did not compile | 3 | 68.1 | 143k |
| p2m-2499fdA02 | QAlgorithms.Papers.Independent.Intakethird100.P20.regularized_base_is_quasipointLemma 3.4 | Opentried before in trial 1 | Not proved: did not compile | 3 | 63.5 | 162k |
| p2m-12ea27A03 | QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialProposition 3.1 | Opentried before in trial 1 | Not proved: a sorry left | 3 | 60.1 | 323k |
| p2m-f10c9dA04 | SongZipkinFluct.Monotone.corollary2Corollary 2: the finite-horizon basestock levels y^*_n(i) converge as n\to\infty | Opentried before in trial 1 | Not proved: a sorry left | 3 | 63.4 | 196k |
Trial 3 · 9 to 12 November 2026 · 3 of 11 proved
A pre-written skeleton and the remaining goals in each fix message. 3 of 11 proved, among them O04, the first open leaf proved in any trial, which prove2me then accepted. O04 turned out short, and nobody had tried it: an existence proof, not a rate. About two thirds of the trial's 58 hours were lost to outages; no leaf lost model time.
Budget per leaf: 45 / 45 for open leaves (30 / 30 for calibration), then 2 fix rounds of 20 / 20.
| Leaf | Theorem | Kind | Result | Rounds | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-a146ceF01 | CodedSorting.proposition_7_1_transposeProposition 7.1 (rearrangement part; companion: proposition_7_1_correctness) — \pi and \pi^{-1} are three adjacent field interchanges | Open | Not proved: a sorry left | 3 | 79.3 | 270k |
| p2m-c3c3c4F02 | APSCounting893.Main.generic_linear_noether_normalizationLemma 3.15 — Generic linear Noether normalization | Open | Not proved: a sorry left | 3 | 78.0 | 516k |
| p2m-8503e6F03 | BirthdayEntropy.Soundness.pair_avoidanceLemma 4.1 — Pair avoidance under fixed-size sampling | Open | Not proved: a sorry left | 3 | 69.8 | 209k |
| p2m-c0eca8O01 | HolevoCQ.decoder_subPOVM§4: the decision operators (11) satisfy \sum_uX_u\le I | Open | Not proved: did not compile | 3 | 78.5 | 321k |
| p2m-83aae8O02 | BBPSSW.Distill.typical_set_sqrtTypical sets of size 2^{H+O(\sqrt n)} carry almost all the weight | Open | Not proved: a sorry left | 3 | 94.5 | 874k |
| p2m-d1f097O03 | DigitalPrivacy.DataStrategy.proof_thm1_case_psi_lt_PrhoProof of Theorem 1, case \psi<P\rho (corrected: \psi>0): C_{{inv}}(\xi)<C_h(\xi) for every \xi>0 | Open | Not proved: did not compile | 3 | 80.6 | 233k |
| p2m-7f145aO04 | SeqPostedPricing.MatroidSPM.accept_prob_eq_quantile§4: offered the price F_i^{-1}(1-q_i), agent i accepts with probability exactly q_i | Open | Proved in round 2 | 2 | 53.9 | 132k |
| p2m-3ad60fC07 | WilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4A finite maximal drawable supergraph exists | Calibration | Proved in round 3 | 3 | 63.6 | 231k |
| p2m-eb3303C08 | AvramDividend.Classical.avram_ubv_finite9_uniformIntegrable_one_of_common_integrable_dominatorAny uniformly dominated L1 family is uniformly integrable at exponent one | Calibration | Not proved: did not compile | 3 | 42.4 | 175k |
| p2m-2b384dC09 | FiniteEnergyPlane.integral_norm_sq_le_of_skewPoincaré inequality on an interval for a curve vanishing at the left endpoint, with a skew-twisted derivative | Calibration | Not proved: a sorry left | 3 | 57.5 | 242k |
| p2m-479463C10 | BookProof.ChapterF3.dressed_vacuum_bessel`BookProof.ChapterF3.dressed_vacuum_bessel` {ι : Type*} (x : ι → E') (hx : Orthonormal ℂ x) (v : E') (hv : ‖v‖ = 1) (s : Finset ι) : ∑ j ∈ s, ‖inner (𝕜 | Calibration | Proved | 1 | 9.0 | 63k |
Trial 4 · 12 to 13 November 2026 · 1 of 8 proved
One sorry at a time, and a stop when a fix round changes nothing. Leaves now run as tasks on the gateway's queue. 1 of 8 proved (calibration leaf C11). Six leaves that earlier trials had failed were tried again: none proved, and no hole closed. Every hard first round ran into its 45-minute cap, half of the model time in long reasoning turns.
Budget per leaf: 45 / 45, then 2 fix rounds of 20 / 20.
| Leaf | Theorem | Kind | Result | Rounds | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-63eff5C11 | Ideal.Factors.quotientResidueEquiv_mkFactors: quotient Residue Equiv mk | Calibration | Proved | 1 | 3.7 | 21k |
| p2m-6f7dceC12 | InsiderDonations.price_recast§3.2, p.9 — P^*=\eta v+[1-\eta]\bar v-\lambda^*u-\sum_i(\psi_i-\bar\psi)/(2\tau_o) | Calibration | Not scoreda fix round ended with no tool call (a tooling fault, since fixed), so the leaf lost its last round | 2 | 50.4 | 159k |
| p2m-12ea27A03 | QAlgorithms.Papers.Independent.Intakethird100.P20.complete_quantum_presheaf_trivialProposition 3.1 | Opentried before in trials 1 and 2 | Not proved: stopped, no progress | 2 | 65.4 | 311k |
| p2m-f10c9dA04 | SongZipkinFluct.Monotone.corollary2Corollary 2: the finite-horizon basestock levels y^*_n(i) converge as n\to\infty | Opentried before in trials 1 and 2 | Not proved: a sorry left | 3 | 83.0 | 232k |
| p2m-a146ceF01 | CodedSorting.proposition_7_1_transposeProposition 7.1 (rearrangement part; companion: proposition_7_1_correctness) — \pi and \pi^{-1} are three adjacent field interchanges | Opentried before in trial 3 | Not proved: a sorry left | 3 | 81.1 | 666k |
| p2m-c0eca8O01 | HolevoCQ.decoder_subPOVM§4: the decision operators (11) satisfy \sum_uX_u\le I | Opentried before in trial 3 | Not proved: stopped, no progress | 2 | 65.2 | 348k |
| p2m-83aae8O02 | BBPSSW.Distill.typical_set_sqrtTypical sets of size 2^{H+O(\sqrt n)} carry almost all the weight | Opentried before in trial 3 | Not proved: a sorry left | 3 | 85.3 | 268k |
| p2m-d1f097O03 | DigitalPrivacy.DataStrategy.proof_thm1_case_psi_lt_PrhoProof of Theorem 1, case \psi<P\rho (corrected: \psi>0): C_{{inv}}(\xi)<C_h(\xi) for every \xi>0 | Opentried before in trial 3 | Not proved: a sorry left | 3 | 85.2 | 249k |
Trial 5 · 13 to 14 November 2026 · 2 of 3 proved
The model first writes an outline of the proof in which every step it has not proved is a sorry hole. Each hole then becomes a task of its own, after a quick pass of standard tactics, a ledger keeps every hole Lean has checked, and an assembly puts the pieces together. 2 of 3 leaves proved. F01, an open leaf that trials 3 and 4 had failed, was proved and accepted by prove2me: its first outline left the main claim in two holes, which failed twice; a second outline split the index into four fields, as the leaf’s own statement does, and all 6 of its holes were proved and assembled on the first try. Calibration leaf C12 was proved on the second try of its assembly. We stopped O02, which is not scored.
Budget per leaf: 90 steps / 90 min for the outline and for each hole, then 2 fix rounds of 20 / 20; up to 2 tries per task.
| Leaf | Theorem | Kind | Result | Tasks | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-a146ceF01 | CodedSorting.proposition_7_1_transposeProposition 7.1 (rearrangement part; companion: proposition_7_1_correctness) — \pi and \pi^{-1} are three adjacent field interchanges | Opentried before in trials 3 and 4 | Proved with outline 2 | 192 outlines | 1208.8 | 3.6M |
| p2m-6f7dceC12 | InsiderDonations.price_recast§3.2, p.9 — P^*=\eta v+[1-\eta]\bar v-\lambda^*u-\sum_i(\psi_i-\bar\psi)/(2\tau_o) | Calibrationtried before in trial 4 | Proved | 41 outline | 135.3 | 468k |
| p2m-83aae8O02 | BBPSSW.Distill.typical_set_sqrtTypical sets of size 2^{H+O(\sqrt n)} carry almost all the weight | Opentried before in trials 3 and 4 | Not scoredstopped by us: its first outline proved none of its 6 holes, and a harness bug (since fixed) left 11 of its second outline’s 12 holes in place | 102 outlines | 641.9 | 1.9M |
Each leaf’s outlines, newest first, with what became of every hole; under each, every task (outline, hole, try and assembly) with its counts, as in “Working on now”.
p2m-a146ce (F01)
Proved: outline 2 assembled
Outline 2proved (assembled)written in 27 steps · 33.1 min
- Hole 1proved
- Hole 2proved
- Hole 3proved in try 2 (try 1: did not compile)
- Hole 4proved
- Hole 5proved
- Hole 6proved
- Assemblyproved: our full check passed
Outline 1abandoned: holes 6 and 7 not proved in 2 trieswritten in 29 steps · 25.4 min
- Hole 1proved
- Hole 2proved
- Hole 3proved
- Hole 4proved
- Hole 5proved
- Hole 6not proved in 2 tries (did not compile, then did not compile)
- Hole 7not proved in 2 tries (did not compile, then a sorry left)
p2m-a146ce (F01): every task (19)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 29 | 25.4 | 135k | 1.2M |
| Hole 1pre-pass timed out | 1 | 1 | Proved | 25 | 19.5 | 80k | 823k |
| Hole 2pre-pass did not close it | 1 | 1 | Proved | 35 | 38.8 | 235k | 1.3M |
| Hole 3pre-pass did not close it | 1 | 1 | Proved | 32 | 19.6 | 104k | 880k |
| Hole 4pre-pass did not close it | 1 | 1 | Proved | 81 | 93.9 | 303k | 5.1M |
| Hole 5pre-pass did not close it | 1 | 1 | Proved | 20 | 6.3 | 29k | 332k |
| Hole 6pre-pass did not close it | 1 | 1 | Not proved: did not compile | 92 | 130.2 | 364k | 5.4M |
| Hole 6 | 1 | 2 | Not proved: did not compile | 65 | 130.6 | 363k | 3.4M |
| Hole 7pre-pass did not close it | 1 | 1 | Not proved: did not compile | 45 | 130.3 | 335k | 3.2M |
| Hole 7 | 1 | 2 | Not proved: a sorry left | 105 | 130.4 | 278k | 6.3M |
| Outline | 2 | 1 | Outline written | 27 | 33.1 | 87k | 1.3M |
| Hole 1pre-pass did not close it | 2 | 1 | Proved | 31 | 39.0 | 102k | 1.4M |
| Hole 2pre-pass did not close it | 2 | 1 | Proved | 36 | 55.8 | 141k | 2.4M |
| Hole 3pre-pass did not close it | 2 | 1 | Not proved: did not compile | 89 | 130.3 | 428k | 5.9M |
| Hole 3 | 2 | 2 | Proved | 38 | 90.1 | 245k | 2.0M |
| Hole 4pre-pass did not close it | 2 | 1 | Proved | 38 | 66.7 | 166k | 2.8M |
| Hole 5pre-pass did not close it | 2 | 1 | Proved | 18 | 13.5 | 38k | 515k |
| Hole 6pre-pass did not close it | 2 | 1 | Proved | 39 | 55.3 | 218k | 2.2M |
| Assembly | 2 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 19 tasks | 845 | 1208.8 | 3.6M | 46.8M |
Outline 1: 7 holes, all extracted as lemmas; the pre-pass of standard tactics closed 0 of 7 (709 s).
Outline 2: 6 holes, all extracted as lemmas; the pre-pass of standard tactics closed 0 of 6 (470 s).
p2m-6f7dce (C12)
Proved: outline 1 assembled
Outline 1proved (assembled)written in 66 steps · 110.2 min
- Hole 1proved
- Assemblyproved: our full check passed (try 2)
p2m-6f7dce (C12): every task (4)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 66 | 110.2 | 371k | 3.9M |
| Hole 1pre-pass did not close it | 1 | 1 | Proved | 25 | 25.1 | 97k | 917k |
| Assembly | 1 | 1 | Assembly did not check | 0 | 0.0 | 0 | 0 |
| Assembly | 1 | 2 | Proved | 0 | 0.0 | 0 | 0 |
| All 4 tasks | 91 | 135.3 | 468k | 4.8M |
Outline 1: 1 hole, extracted as a lemma; the pre-pass of standard tactics closed 0 of 1 (71 s).
p2m-83aae8 (O02)
Stopped: outline 2 was abandoned (its tasks were cancelled)
Outline 2abandoned: its tasks were cancelledwritten in 44 steps · 61.0 min
- Hole 1not tried (outline abandoned)
- In placeholes 2 to 12, not tried (outline abandoned)
Outline 1abandoned: holes 1, 2 and 3 and the holes in place not provedwritten in 32 steps · 55.8 min
- Hole 1not proved in 2 tries (a fault on our side, then stopped, no progress)
- Hole 2not proved in 2 tries (stopped, no progress, then a sorry left)
- Hole 3not proved: a sorry left
- In placeholes 4 to 6, not proved: stopped, no progress
p2m-83aae8 (O02): every task (10)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 32 | 55.8 | 115k | 1.8M |
| Hole 1pre-pass did not close it | 1 | 1 | Not proved: a fault on our side | 18 | 18.9 | 77k | 428k |
| Hole 1 | 1 | 2 | Not proved: stopped, no progress | 91 | 110.0 | 300k | 5.9M |
| Hole 2pre-pass did not close it | 1 | 1 | Not proved: stopped, no progress | 25 | 110.2 | 260k | 1.1M |
| Hole 2 | 1 | 2 | Not proved: a sorry left | 127 | 87.6 | 268k | 7.5M |
| Hole 3pre-pass did not close it | 1 | 1 | Not proved: a sorry left | 82 | 130.4 | 403k | 5.4M |
| In placethe 3 holes left in place, in one task | 1 | 1 | Not proved: stopped, no progress | 100 | 68.0 | 335k | 6.2M |
| Outline | 2 | 1 | Outline written | 44 | 61.0 | 162k | 2.5M |
| Hole 1pre-pass timed out | 2 | 1 | Cancelled | – | – | – | – |
| In placethe 11 holes left in place, in one task | 2 | 1 | Cancelled | – | – | – | – |
| All 10 tasks | 519 | 641.9 | 1.9M | 30.7M |
Outline 1: 6 holes, 3 extracted as lemmas, 3 left in place; the pre-pass of standard tactics closed 0 of 3 (268 s).
Outline 2: 12 holes, 1 extracted as a lemma, 11 left in place; the pre-pass of standard tactics closed 0 of 1 (118 s).
Trial 6 · 15 November 2026 · 2 of 2 proved
Trial 6 ran on 15 November 2026 on scaffold v7: 2 leaves (1 calibration, 1 open), 2 of 2 runs proved; prove2me accepted 1 proof from it and 2 leaves closed by reuse.
The outline task now starts with a plan, every step of it becomes a sorry hole (at most 12 holes and 300 lines), and an outline that does not compile is repaired by turning the parts Lean rejects into holes. All 3 leaves were closed. Open leaf p2m-dbd69f was proved by the model and accepted by prove2me: its first outline had a false hole, which the model disproved; its second outline, the paper's duality argument in four steps, had every hole proved and was assembled. Open leaf p2m-f181df was closed by reuse of a theorem already proved on prove2me, and calibration leaf p2m-3ad60f was proved by assembly.
Budget per leaf: one round of 110 steps / 110 min for the outline and for each hole, then one fix round of 20 / 20; up to 2 tries per task.
| Leaf | Theorem | Kind | Result | Tasks | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-3ad60fC07 | WilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4A finite maximal drawable supergraph exists | Calibrationtried before in trial 3 | ProvedProved: outline 1 assembled | 31 outline | 8.3 | 69k |
| p2m-dbd69fN03 | PerturbedWF.eq_2_3Eq. (2.3): \langle v,x-y\rangle\le\|x-y\| for v\in B^* and x,y\in\mathcal C | Open | Proved with outline 2Proved: outline 2 assembled | 102 outlines | 165.6 | 695k |
Each leaf’s outlines, newest first, with what became of every hole; under each, every task (outline, hole, try and assembly) with its counts, as in “Working on now”.
p2m-3ad60f (C07)
Proved: outline 1 assembled
Outline 1proved (assembled)written in 19 steps · 6.6 min
- Hole 1closed by the index (reuse)
- Hole 2proved
- Hole 3closed by the pre-pass
- Assemblyproved: our full check passed
p2m-3ad60f (C07): every task (3)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 19 | 6.6 | 55k | 440k |
| Hole 2pre-pass did not close it | 1 | 1 | Proved | 6 | 1.7 | 14k | 86k |
| Assembly | 1 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 3 tasks | 25 | 8.3 | 69k | 526k |
Outline 1: 3 holes, all extracted as lemmas; the pre-pass of standard tactics closed 2 of 3 (479 s).
p2m-dbd69f (N03)
Proved: outline 2 assembled
Outline 2proved (assembled)written in 28 steps · 12.8 min
- Hole 1proved
- Hole 2in place, proved (checked in the assembly)
- Hole 3proved
- Hole 4proved
- Assemblyproved: our full check passed
Outline 1abandoned: its tasks were cancelledwritten in 59 steps · 27.0 min
- Hole 1in place, not tried (outline abandoned)
- Hole 2try 1: our check failed; try 2 not tried (outline abandoned)
p2m-dbd69f (N03): every task (10)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 59 | 27.0 | 106k | 2.6M |
| Hole 2pre-pass did not close it | 1 | 1 | Our check failed | 19 | 7.0 | 39k | 386k |
| Hole 2 | 1 | 2 | Cancelled | – | – | – | – |
| In placethe 1 hole left in place, in one task | 1 | 1 | Cancelled | – | – | – | – |
| Outline | 2 | 1 | Outline written | 28 | 12.8 | 50k | 1.1M |
| Hole 1pre-pass did not close it | 2 | 1 | Proved | 21 | 4.8 | 21k | 362k |
| Hole 3pre-pass did not close it | 2 | 1 | Proved | 47 | 44.2 | 176k | 2.4M |
| Hole 4pre-pass did not close it | 2 | 1 | Proved | 15 | 7.6 | 28k | 341k |
| In placethe 1 hole left in place, in one task | 2 | 1 | Filled in, checked by the assembly | 118 | 62.2 | 274k | 9.2M |
| Assembly | 2 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 10 tasks | 307 | 165.6 | 695k | 16.3M |
Outline 1: 2 holes, 1 extracted as a lemma, 1 left in place; the pre-pass of standard tactics closed 0 of 1 (89 s).
Outline 2: 4 holes, 3 extracted as lemmas, 1 left in place; the pre-pass of standard tactics closed 0 of 3 (316 s).
Trial 7 · 15 to 17 November 2026 · 3 of 4 proved
Trial 7 ran from 15 to 17 November 2026 on scaffold v8: 4 leaves (1 calibration, 3 open), 3 of 4 runs proved; prove2me accepted 2 proofs from it.
Scaffold v8: v7 with what trial 6 taught. A counterexample to a false hole goes into the next outline's prompt, and an outline that proves the leaf ends it. 3 of 4 leaves were proved, and prove2me accepted both open leaves among them. Open leaf p2m-aec6a2 was proved from its first outline: its two holes were closed and the proof assembled. Open leaf p2m-d00cfc took three outlines: standard tactics closed every hole of the first two, which the outline check then refused as having no real step (since changed: such an outline is a proof); the third had 2 holes, both proved, and was assembled. Calibration leaf p2m-3ad60f was proved again, through an outline with one hole. Open leaf p2m-115f6c (Pinsker's inequality, the hard one) was not proved: of its three outlines, with 8, 3 and 9 holes, most holes were proved, but in each some hole failed both tries or was shown to be false.
Budget per leaf: one round of 110 steps / 110 min for the outline and for each hole, then one fix round of 20 / 20; up to 2 tries per task.
| Leaf | Theorem | Kind | Result | Tasks | Model min | Tokens computed |
|---|---|---|---|---|---|---|
| p2m-3ad60fC07 | WilliamsonShmoys.planar_drawing_maximal_supergraph__p2m_97829890_428d_4657_a899_f7e200a68a7c_a4A finite maximal drawable supergraph exists | Calibrationtried before in trials 3 and 6 | ProvedProved: outline 1 assembled | 31 outline | 71.4 | 202k |
| p2m-aec6a2N10 | thm_5_6Chapter 5, thm_5_6 | Open | ProvedProved: outline 1 assembled | 41 outline | 20.9 | 138k |
| p2m-d00cfcN07 | MatchIndiff.PublicSM.lemma_2_1Lemma 2.1 — accepted profiles of a fixed board form a product, and a row and board fix the decoded component | Open | Proved with outline 3Proved: outline 3 assembled | 63 outlines | 47.3 | 270k |
| p2m-115f6cN08 | QAlgorithms.Papers.Independent.Intake100.P73.pinsker_full_l1Pinsker inequality in the source full-L1/nats convention | Open | Not provedNot proved: outline 3 was refuted (hole 9 is false) | 273 outlines | 1267.6 | 5.8M |
Each leaf’s outlines, newest first, with what became of every hole; under each, every task (outline, hole, try and assembly) with its counts, as in “Working on now”.
p2m-3ad60f (C07)
Proved: outline 1 assembled
Outline 1proved (assembled)written in 73 steps · 62.8 min
- Hole 1proved
- Assemblyproved: our full check passed
p2m-3ad60f (C07): every task (3)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 73 | 62.8 | 157k | 4.2M |
| Hole 1pre-pass did not close it | 1 | 1 | Proved | 24 | 8.6 | 45k | 741k |
| Assembly | 1 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 3 tasks | 97 | 71.4 | 202k | 4.9M |
Outline 1: 1 hole, extracted as a lemma; the pre-pass of standard tactics closed 0 of 1 (91 s).
p2m-aec6a2 (N10)
Proved: outline 1 assembled
Outline 1proved (assembled)written in 26 steps · 6.9 min
- Hole 1proved
- Hole 2in place, proved (checked in the assembly)
- Assemblyproved: our full check passed
p2m-aec6a2 (N10): every task (4)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 26 | 6.9 | 68k | 676k |
| Hole 1pre-pass did not close it | 1 | 1 | Proved | 22 | 8.0 | 37k | 429k |
| In placethe 1 hole left in place, in one task | 1 | 1 | Filled in, checked by the assembly | 24 | 6.0 | 34k | 526k |
| Assembly | 1 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 4 tasks | 72 | 20.9 | 138k | 1.6M |
Outline 1: 2 holes, 1 extracted as a lemma, 1 left in place; the pre-pass of standard tactics closed 0 of 1 (83 s).
p2m-d00cfc (N07)
Proved: outline 3 assembled
Outline 3proved (assembled)written in 17 steps · 9.2 min
- Hole 1proved
- Hole 2proved
- Assemblyproved: our full check passed
Outline 2refuted: degenerate (a hole is the leaf’s own claim)written in 30 steps · 22.3 min
Outline 1refuted: degenerate (a hole is the leaf’s own claim)written in 23 steps · 12.5 min
p2m-d00cfc (N07): every task (6)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Ended: degenerate outline | 23 | 12.5 | 76k | 687k |
| Outline | 2 | 1 | Ended: degenerate outline | 30 | 22.3 | 103k | 1.3M |
| Outline | 3 | 1 | Outline written | 17 | 9.2 | 64k | 551k |
| Hole 1pre-pass did not close it | 3 | 1 | Proved | 9 | 2.1 | 15k | 129k |
| Hole 2pre-pass did not close it | 3 | 1 | Proved | 7 | 1.2 | 13k | 93k |
| Assembly | 3 | 1 | Proved | 0 | 0.0 | 0 | 0 |
| All 6 tasks | 86 | 47.3 | 270k | 2.7M |
Outline 3: 2 holes, all extracted as lemmas; the pre-pass of standard tactics closed 0 of 2 (192 s).
p2m-115f6c (N08)
Not proved: outline 3 was refuted (hole 9 is false)
Outline 3refuted: hole 9 is falsewritten in 99 steps · 66.1 min
- Hole 1proved
- Hole 2not proved in 2 tries (a sorry left, then a sorry left)
- Hole 3proved
- Hole 4proved
- Hole 5proved
- Hole 6proved
- Hole 7proved
- Hole 8try 1: a sorry left; try 2 not tried (outline refuted)
- Hole 9false, disproved by the model
Outline 2abandoned: hole 3 not proved in 2 trieswritten in 57 steps · 72.3 min
- Hole 1proved
- Hole 2proved
- Hole 3not proved in 2 tries (a sorry left, then a sorry left)
Outline 1refuted: holes 6 and 8 are falsewritten in 36 steps · 19.4 min
- Hole 1closed by the pre-pass
- Hole 2proved
- Hole 3proved
- Hole 4proved
- Hole 5proved
- Hole 6false, disproved by the model in try 2 (try 1: a sorry left)
- Hole 7not proved in 2 tries (did not compile, then a sorry left)
- Hole 8false, disproved by the model
p2m-115f6c (N08): every task (27)
| Task | Outline | Try | Result | Steps | Model min | Tokens computed | With cache re-reads |
|---|---|---|---|---|---|---|---|
| Outline | 1 | 1 | Outline written | 36 | 19.4 | 106k | 2.1M |
| Hole 2pre-pass did not close it | 1 | 1 | Proved | 11 | 3.2 | 14k | 144k |
| Hole 3pre-pass did not close it | 1 | 1 | Proved | 20 | 5.0 | 49k | 519k |
| Hole 4pre-pass did not close it | 1 | 1 | Proved | 30 | 8.6 | 47k | 768k |
| Hole 5pre-pass did not close it | 1 | 1 | Proved | 31 | 13.4 | 69k | 1.1M |
| Hole 6pre-pass did not close it | 1 | 1 | Not proved: a sorry left | 128 | 48.1 | 234k | 8.8M |
| Hole 6 | 1 | 2 | Disproved: the hole is false | 97 | 60.0 | 298k | 7.5M |
| Hole 7pre-pass did not close it | 1 | 1 | Not proved: did not compile | 128 | 67.0 | 301k | 9.2M |
| Hole 7 | 1 | 2 | Not proved: a sorry left | 112 | 95.4 | 527k | 10.3M |
| Hole 8pre-pass did not close it | 1 | 1 | Disproved: the hole is false | 128 | 49.7 | 258k | 9.8M |
| Outline | 2 | 1 | Outline written | 57 | 72.3 | 241k | 5.0M |
| Hole 1pre-pass did not close it | 2 | 1 | Proved | 38 | 28.2 | 115k | 1.7M |
| Hole 2pre-pass did not close it | 2 | 1 | Proved | 21 | 14.8 | 50k | 480k |
| Hole 3pre-pass did not close it | 2 | 1 | Not proved: a sorry left | 106 | 116.2 | 624k | 8.5M |
| Hole 3 | 2 | 2 | Not proved: a sorry left | 80 | 129.2 | 666k | 7.9M |
| Outline | 3 | 1 | Outline written | 99 | 66.1 | 241k | 6.4M |
| Hole 1pre-pass did not close it | 3 | 1 | Proved | 24 | 4.4 | 52k | 643k |
| Hole 2pre-pass did not close it | 3 | 1 | Not proved: a sorry left | 40 | 130.2 | 517k | 4.6M |
| Hole 2 | 3 | 2 | Not proved: a sorry left | 119 | 109.7 | 395k | 9.1M |
| Hole 3pre-pass did not close it | 3 | 1 | Proved | 31 | 29.9 | 125k | 1.5M |
| Hole 4pre-pass did not close it | 3 | 1 | Proved | 66 | 26.9 | 119k | 3.8M |
| Hole 5pre-pass did not close it | 3 | 1 | Proved | 65 | 26.6 | 90k | 2.5M |
| Hole 6pre-pass did not close it | 3 | 1 | Proved | 104 | 44.0 | 200k | 6.0M |
| Hole 7pre-pass did not close it | 3 | 1 | Proved | 74 | 33.5 | 135k | 3.3M |
| Hole 8pre-pass did not close it | 3 | 1 | Not proved: a sorry left | 19 | 59.5 | 265k | 1.1M |
| Hole 8 | 3 | 2 | Cancelled | – | – | – | – |
| Hole 9pre-pass did not close it | 3 | 1 | Disproved: the hole is false | 13 | 6.3 | 39k | 409k |
| All 27 tasks | 1677 | 1267.6 | 5.8M | 113.2M |
Outline 1: 8 holes, all extracted as lemmas; the pre-pass of standard tactics closed 1 of 8 (840 s).
Outline 2: 3 holes, all extracted as lemmas; the pre-pass of standard tactics closed 0 of 3 (262 s).
Outline 3: 9 holes, all extracted as lemmas; the pre-pass of standard tactics closed 0 of 9 (300 s).
How to read this
- The Lean kernel is the judge. A run counts as proved only when our full check passes (the proof builds, its theorem is exactly the leaf’s, and it uses no
sorryand only the axiomspropext,Classical.choiceandQuot.sound), with Lean v4.33.1 and Mathlib0df444aas on prove2me. A submission counts as accepted only when prove2me’s verdict says so. We do not grade the model’s work ourselves. - Calibration and open leaves. A calibration leaf had already been proved by someone else when we chose it. It tells us whether the setup works; the model is not shown that proof, and we never submit one. An open leaf had no accepted proof when we chose it. An open leaf “closed by reuse” has the same statement as a theorem already proved on prove2me, or is a part of one: we submit the one-line proof that uses it, name its author, and do not count it as proved by the model.
- Every run counts. A run is one leaf in one trial: a first round, and from trial 2 up to two fix rounds in which the model gets Lean’s errors. From trial 5 a run is all of a leaf’s tasks: its outlines, one task per hole, the tries and the assembly. Failed runs stay in the tables. A leaf tried again in a later trial counts as one more run, and once among the open leaves tried. Runs cut short by a fault on our side, not the model’s, or stopped by us because of one, say “Not scored” and why (4 so far); they still count among the runs.
- Model minutes are the time the model spent generating, summed over a run’s rounds (from trial 5, its tasks). Tokens computed are the work the GPU did: what the model read that was not already in the cache, plus what it wrote. An agent reads the same files again and again, and the cache serves those re-reads: counted with them, the tokens are about 13 times as many (92% of everything read and written were re-reads). Hover a token count for both numbers. Two first tries in trial 2 that the gateway’s personal-data filter cut off are not in the counts.
- Working on now and the trials from trial 5 show each leaf’s outlines, newest first. An outline is in progress, refuted (one of its holes is false, or its shape was refused), proved (its holes put together passed our full check) or abandoned for a new outline. Each hole is proved by its own task, false (the model disproved it), closed before any task by the pre-pass of standard tactics, by our index of prove2me’s library (reuse) or by the ledger (proved before), or it is running, queued, not tried or cancelled; the holes left in the outline’s file (“in place”) share one task. A cancelled task says why: its outline refuted, replaced or abandoned after a hole failed, or its leaf proved. A leaf’s tokens count every task, each outline and each try included: tokens computed as above, and the count with the cache re-reads beside them. Open “Every task” for each task’s result, steps, model minutes and tokens. A running task’s steps and tokens show as the gateway counts them where it can, else once the task ends.
- Small samples. 48 runs in 7 trials, each with a different scaffold, budget and set of leaves. They show what happened, not a success rate.
- These are our own measurements, from the records of each run, not estimates. The model runs on the AIs@home network, which today is mostly the maintainer’s own GPUs, some of them rented. The data is published as JSON:
/science/prove2me.json.
What comes next
- Trial 5 asked whether one task per hole closes holes that whole-leaf rounds could not. On F01 it did, once; on O02 it closed none before we stopped it.
- Trial 8, above: whether its scaffold, described under “Working on now”, lets the model close easier open leaves, chosen by quick outline probes.
- Then a random sample of open leaves at a fixed budget, with the pass line written down before the run: the test of whether this works at a useful rate, and what a proof costs.
- The question behind it: can an open model on one consumer GPU close open leaves often enough that volunteers’ GPUs could do real work here?