{"record": {"aliases": ["Pythagoras-Prover: Advancing Efficient Formal Proving via Augmented Lean Formalisation"], "artifacts": [{"id": "arxiv:2606.12594", "kind": "paper", "url": "https://arxiv.org/abs/2606.12594"}], "authors": ["Joshua Ong Jun Leang", "Zheng Zhao", "Mihaela Cătălina Stoian", "Qiyuan Xu", "Haonan Li", "Wenda Li", "Shay B. Cohen", "Eleonora Giunchiglia"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2606.12594", "source_url": "https://arxiv.org/abs/2606.12594"}, "categories": ["Language & Knowledge", "Mathematics & Formal Science", "cs.AI"], "description": {"en": "A released, contamination-sensitive benchmark of augmented Lean formal statements for evaluating formal theorem provers."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2606.12594", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2606.12594", "title": "MiniF2F-ALF"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2606.12594", "modality": null, "name": "MiniF2F-ALF", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-09-25T04:40:22Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_pythagoras-prover-advancing-efficient-form_8a18cd25", "origin_source": "arxiv", "recognition_confidence": "0.6", "record_sha256": "8d8afa62bf06b8429740c545d4fc3e6d6ffeb8e231c32e217481400e77e1aac5", "relation": "introduces", "review_model": "codex-cli-default", "review_state": "ai-reviewed", "reviewed_at": "2026-09-24T18:26:47.670188Z", "source_url": "https://arxiv.org/abs/2606.12594"}, "publisher": null, "released": "2026-06-10", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2606.12594", "source_url": "https://arxiv.org/abs/2606.12594"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2606-12594", "source": "claire_radar", "source_benchmark_id": "2606.12594", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["Pythagoras-Prover: Advancing Efficient Formal Proving via Augmented Lean Formalisation"], "annotation": "Unknown", "applicationDomains": ["Mathematics & Formal Science"], "area": "Language & Knowledge", "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "Efficient formal proving and contamination-sensitive evaluation are likely to receive substantial near-term attention.", "score": 66}, "capabilities": [], "confidence": "Medium", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-24T18:26:47.670188Z", "inputHash": "845185289095e32e23445582fbebe5c1851aa40892e2ac163deb46b3af6a25b7", "model": "codex-cli-default"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "We release MiniF2F-ALF", "canonicalNameSource": "abstract", "decisionReason": "The abstract formally names and releases the benchmark, identifies its formal-statement evaluation material, and uses accuracy for model comparison.", "evaluationEvidence": {"reuse": "We release MiniF2F-ALF", "scoring": "every evaluated model loses accuracy", "task": "variants of formal statements"}, "model": "codex-cli-default", "reviewedAt": "2026-09-24T18:26:47.670188Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "A released, contamination-sensitive benchmark of augmented Lean formal statements for evaluating formal theorem provers.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["benchmark term in abstract", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We release MiniF2F-ALF, an ALF-mutated contamination-sensitive benchmark on which every evaluated model loses accuracy; here our 32B remains strongest and our 4B matches the prior state of the art, Goedel-Prover-V2-32B."}, "familyId": "bmf_c76332523803", "firstSeenAt": "2026-08-19", "heat": null, "id": "bm_pythagoras-prover-advancing-efficient-form_8a18cd25", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2606.12594", "pdf": "https://arxiv.org/pdf/2606.12594", "project": null, "report": "https://arxiv.org/abs/2606.12594"}, "metrics": [], "motivation": "Modern Lean theorem provers achieve strong performance only with substantial training and inference compute, driven in part by scarce verified proof data and the long reasoning traces of formal proof search, making both supervised fine-tuning (SFT) and sampling expensive.", "name": "MiniF2F-ALF", "oneLine": "A released, contamination-sensitive benchmark of augmented Lean formal statements for evaluating formal theorem provers.", "paperTitle": "Pythagoras-Prover: Advancing Efficient Formal Proving via Augmented Lean Formalisation", "primaryDomain": "Mathematics & Formal Science", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://arxiv.org/abs/2606.12594", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-08-19T11:05:13.395391Z"}, "publications": [], "readiness": "Paper only", "recognitionConfidence": 0.6, "relation": "introduces", "releasedAt": "2026-06-10", "reviewLifecycle": {"state": "pending"}, "source": {"authors": ["Joshua Ong Jun Leang", "Zheng Zhao", "Mihaela C\\u{a}t\\u{a}lina Stoian", "Qiyuan Xu", "Haonan Li", "Wenda Li", "Shay B. Cohen", "Eleonora Giunchiglia"], "categories": ["cs.AI"], "id": "2606.12594", "title": "Pythagoras-Prover: Advancing Efficient Formal Proving via Augmented Lean Formalisation", "type": "arxiv", "url": "https://arxiv.org/abs/2606.12594"}, "sourceUpdatedAt": "2026-06-10T18:43:11Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "It tests whether prover performance persists when familiar formal statements are systematically mutated."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
