datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
proofwriter
Dataset Card for "proofwriter"
More Information needed
proof-bundlesproofwriter_processed_OWAcve-proof-corpus
CVE Proof Corpus
Six real vulnerability classes, each with a machine-checkable proof that the shipped fix
eliminates it — and a checker that shares no code with whatever produced the proof.
Every record carries the safety relation, the guard the upstream project shipped, the declared
attacker domain, and the nonnegative multipliers that prove the guard implies safety. All six verify.
pip install "certkit@git+https://github.com/nickharris808/certkit@main"
python verify.py… See the full description on the dataset page: https://huggingface.co/datasets/nickh007/cve-proof-corpus.lean-proofs-v1
Part of the SZL Holdings governed estate — claims are designed to carry checkable receipts. Verification proves integrity & origin, never accuracy or performance.
SZLHOLDINGS/lean-proofs-v1
The complete Lean 4 theorem library for the SZL Holdings Ouroboros Invariant research programme.
Doctrine v10/v11 Canonical Numbers
Metric
Value
Declarations
749
Unique axioms
14 (15 raw, 1 dup)
Sorries
163 (112 baseline + 51 Putnam)… See the full description on the dataset page: https://huggingface.co/datasets/SZLHOLDINGS/lean-proofs-v1.litcoin-proof-of-research
LITCOIN Proof-of-Research Corpus
191,484,662 AI research submissions, produced by 81,224 anonymous contributors and 470 model
variants competing against each other, every row executed in a sandbox and scored.
This is the complete output of the LITCOIN protocol, which ran on Base from March to August 2026.
Autonomous AI agents were paid in a permissionless token to solve real optimization problems across
32 domains. The protocol was discontinued on 20 August 2026. This dataset is… See the full description on the dataset page: https://huggingface.co/datasets/tekkaadan/litcoin-proof-of-research.proofwriter
Dataset Card for "proofwriter"
More Information needed
lean-proof-compression
LeanPolish: Verified Supervision for Lean Proof Compression
A dataset of Lean 4 proof rewrite pairs produced by LeanPolish,
a kernel-verified proof-shortening tool. Every accepted
(original, replacement) pair was kernel-checked under Lean 4.21.0
with Mathlib v4.21.0 before emission, and the rewritten file was
re-elaborated end-to-end by a separate out-of-process verifier.
The dataset is suitable for training models that learn to compress,
simplify, or select proof tactics, and… See the full description on the dataset page: https://huggingface.co/datasets/leanpolish-anon/lean-proof-compression.proofwriter-source
ProofWriter (The Source)
An unmodified copy of AI2's ProofWriter dataset (release V2020.12.3), re-hosted as datasets
configs for convenient loading. The records are faithful to the upstream release — the id-keyed JSON
is preserved as-is; typing and reasoning-graph extraction happen in later stages.
Each config is a {world}-depth-{n} shelf of the synthetic core (OWA/CWA × depths
0/1/2/3/5), split train/dev/test (dev kept as the corpus names it).
Source:… See the full description on the dataset page: https://huggingface.co/datasets/arqa39/proofwriter-source.proofatlas-enriched
ProofAtlas Enriched Dataset v1
ProofAtlas Enriched combines theorem-disjoint Lean retrieval splits with LLM-generated theorem semantic and strategy enrichment. It is designed for premise-retrieval research, theorem-neighborhood retrieval, and qualitative retrieval-evidence analysis.
Source Data
This dataset is derived from erbacher/LeanRank-data, which is distributed under the Apache-2.0 license. The upstream LeanRank data was extracted from mathlib4 using… See the full description on the dataset page: https://huggingface.co/datasets/ZDCSlab/proofatlas-enriched.proofwriter-source
ProofWriter (The Source)
An unmodified copy of AI2's ProofWriter dataset (release V2020.12.3), re-hosted as datasets
configs for convenient loading. The records are faithful to the upstream release — the id-keyed JSON
is preserved as-is; typing and reasoning-graph extraction happen in later stages.
Each config is a {world}-depth-{n} shelf of the synthetic core (OWA/CWA × depths
0/1/2/3/5), split train/dev/test (dev kept as the corpus names it).
Source:… See the full description on the dataset page: https://huggingface.co/datasets/rlhf-and-friends/proofwriter-source.Qwen3.5-9B-EquationalTheories-Proof-eval
Qwen3.5-9B on EquationalTheories-Proof: equational internalization evaluation
Model: the pinned base Qwen3.5-9B (no training)
Evaluation of Qwen/Qwen3.5-9B (revision c202236235762e1c871ad0ccb60c8ee5ba337b9a) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of implication/non-implication relationships between the
4,694 equations of the Equational Theories Project… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-EquationalTheories-Proof-eval.ProofBench
ProofBench Dataset
ProofBench is a comprehensive benchmark dataset for evaluating AI models on mathematical proof generation and verification. The dataset contains competition-level mathematics problems from prestigious international competitions, along with both AI-generated and reference solutions, expert ratings, and detailed marking schemes.
Dataset Overview
Total Problems: 435
Competitions: PUTNAM, IMO, USAMO, EGMO, APMO, TST
Years Covered: 2022-2025
AI Models:… See the full description on the dataset page: https://huggingface.co/datasets/wenjiema02/ProofBench.Qwen3.5-9B-eq70-1m-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-1m on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT (LR 5e-6, two epochs) on 1M loss tokens: 70% WithProof notes (0.70M) + 30% note-conditioned teacher trajectories (0.30M), held-out-first split relationships-v2/ladder v2
Evaluation of Qwen3.5-9B-eq70-1m (revision 1545499d4db014a012dedc6a7b853779ca21d93e) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-1m-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-3m-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-3m on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on 3M loss tokens: 70% notes (2.10M) + 30% trajectories (0.90M)
Evaluation of Qwen3.5-9B-eq70-3m (revision 14fd1fe253524178b6cdfc819ea8d39c4648d496) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of implication/non-implication relationships between the
4… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-3m-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-10m-pc10-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-10m-pc10 on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on the 10M 70/30 mixture plus the 304 study proof chats repeated 10 times (0.62M loss tokens, 5.8%)
Evaluation of Qwen3.5-9B-eq70-10m-pc10 (revision 939e9ffc74c5dfaa08f770fa0c91dbbc1b5b0add) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-10m-pc10-EquationalTheories-Proof-eval.proofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/rlhf-and-friends/proofwriter-mirror.Qwen3.5-9B-eq70-3m-pc10-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-3m-pc10 on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on the 3M 70/30 mixture plus the 304 study proof chats repeated 10 times (0.62M loss tokens, 17%)
Evaluation of Qwen3.5-9B-eq70-3m-pc10 (revision 3752365db8ceda9c519d6f5fa4ef4c9bb26fb604) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-3m-pc10-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-30m-pc10-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-30m-pc10 on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on the 30M 70/30 mixture plus the 304 study proof chats repeated 10 times (0.62M loss tokens, 2.0%)
Evaluation of Qwen3.5-9B-eq70-30m-pc10 (revision c389206d3c8a54a4b4a1c6626effcf1000e3ea21) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-30m-pc10-EquationalTheories-Proof-eval.lean-proof-or-refute-300
Lean Proof-or-Refute 300
Lean Proof-or-Refute 300 is a compact collection of 300 formal reasoning
problems grounded in Lean 4 and Mathlib. Each problem starts from a verified
Mathlib theorem, makes one small numerical or operator mutation, and asks the
model to return either:
a Lean certificate proving the mutated proposition; or
a Lean certificate proving the exact negation of the complete proposition.
The model receives the related source theorem, a bounded source excerpt… See the full description on the dataset page: https://huggingface.co/datasets/xlr8harder/lean-proof-or-refute-300.Qwen3.5-9B-eq70-30m-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-30m on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on 30M loss tokens: 70% notes (21.0M) + 30% trajectories (9.0M)
Evaluation of Qwen3.5-9B-eq70-30m (revision af0ac70cff2cca67261963fa48f7087747defc49) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of implication/non-implication relationships between the
4… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-30m-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-10m-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-10m on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on 10M loss tokens: 70% notes (7.0M) + 30% trajectories (3.0M)
Evaluation of Qwen3.5-9B-eq70-10m (revision 5bfa231a8d33f5e4d5ff49ec40f4cbf2f452ba99) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of implication/non-implication relationships between the
4… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-10m-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-pc16-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-pc16 on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on the 304 user-assistant study proofs only (proof body supervised), repeated 16 times (0.99M loss tokens): the proof-chat control
Evaluation of Qwen3.5-9B-eq70-pc16 (revision ad0dd90c4ab11e4ed8b7b1a1876ae3be718ce77d) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof):… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-pc16-EquationalTheories-Proof-eval.Qwen3.5-9B-eq70-1m-pc10-EquationalTheories-Proof-eval
Qwen3.5-9B-eq70-1m-pc10 on EquationalTheories-Proof: equational internalization evaluation
Model: Qwen3.5-9B full SFT on the 1M 70/30 mixture plus the 304 study proof chats repeated 10 times (0.62M loss tokens, 38% of the mixture)
Evaluation of Qwen3.5-9B-eq70-1m-pc10 (revision f1f6152dba459faec87f14a2ba717267ff96838b) on the equational internalization
protocol (recipe/equational_internalization/EVALUATION_PROTOCOL.md, environment
EquationalTheories-Proof): closed-book recall of… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-1m-pc10-EquationalTheories-Proof-eval.proofjudge
ProofJudge
Paired Lean 4 / Mathlib human-authored proofs: the state a declaration was in when a pull request was
opened, and the state it was in when Mathlib merged it. A code-reviewing judge agent is aligned on a pair when it scores the merged version above the initial one.
Pairs were selected from Mathlib PRs by two filters: a mechanical one (at least 3 changed tokens and at least 5% change relative to the shorter proof) and an LLM judge (claude-sonnet-4-20250514) classifying… See the full description on the dataset page: https://huggingface.co/datasets/SJCaldwell/proofjudge.lm-eval-results-Contamination-contaminated_proof_7b_v1.0_safetensor-private
Dataset Card for Evaluation run of Contamination/contaminated_proof_7b_v1.0_safetensor
Dataset automatically created during the evaluation run of model Contamination/contaminated_proof_7b_v1.0_safetensor
The dataset is composed of 62 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 4 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train"… See the full description on the dataset page: https://huggingface.co/datasets/nyu-dice-lab/lm-eval-results-Contamination-contaminated_proof_7b_v1.0_safetensor-private.proofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/alexdeath53/proofwriter-mirror.Qwen3.5-9B-eq70-3m-v2-EquationalTheories-Proof-nl-eval
Qwen3.5-9B 3m: Equational Theories prose-proof evaluation
Accuracy: 44.85% (296/660 correct).
Single-turn natural-language proof correctness, graded by GPT-5.6-sol. These are model-judged results; generated proofs were not checked by the Lean kernel.
Model
Accuracy
Correct / total
Change vs base
Base
42.73%
282/660
+0.00 pp
1M
38.64%
255/660
-4.09 pp
3M (this dataset)
44.85%
296/660
+2.12 pp
10M
43.64%
288/660
+0.91 pp
The default dataset viewer now shows… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-3m-v2-EquationalTheories-Proof-nl-eval.Qwen3.5-9B-eq70-1m-v2-EquationalTheories-Proof-nl-eval
Qwen3.5-9B 1m: Equational Theories prose-proof evaluation
Accuracy: 38.64% (255/660 correct).
Single-turn natural-language proof correctness, graded by GPT-5.6-sol. These are model-judged results; generated proofs were not checked by the Lean kernel.
Model
Accuracy
Correct / total
Change vs base
Base
42.73%
282/660
+0.00 pp
1M (this dataset)
38.64%
255/660
-4.09 pp
3M
44.85%
296/660
+2.12 pp
10M
43.64%
288/660
+0.91 pp
The default dataset viewer now shows… See the full description on the dataset page: https://huggingface.co/datasets/violetxi/Qwen3.5-9B-eq70-1m-v2-EquationalTheories-Proof-nl-eval.proofwriter-mirror
ProofWriter (The Mirror)
A typed, content-faithful mirror of AI2's ProofWriter dataset (release V2020.12.3),
derived from proofwriter-source. The JSON encoding is cleaned up: the id-keyed dicts
(triple1, Q3, …) become lists of structs that keep their id, every atom representation
is parsed into a typed {subject, relation, object, polarity} triple, and the closed enums
(answer, strategy) are typed. The content stays faithful — nothing renamed, no rows
dropped — and the recursive… See the full description on the dataset page: https://huggingface.co/datasets/arqa39/proofwriter-mirror.