{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:CLESLWG5UDG3G4I66G6NGHWQO7","short_pith_number":"pith:CLESLWG5","schema_version":"1.0","canonical_sha256":"12c925d8dda0cdb3711ef1bcd31ed077efd118a45f01956d83e6e59c586e3038","source":{"kind":"arxiv","id":"2305.12524","version":3},"attestation_state":"computed","paper":{"title":"TheoremQA: A Theorem-driven Question Answering dataset","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jianyu Xu, Max Ku, Ming Yin, Pan Lu, Tony Xia, Wenhu Chen, Xinyi Wang, Xueguang Ma, Yixin Wan","submitted_at":"2023-05-21T17:51:35Z","abstract_excerpt":"The recent LLMs like GPT-4 and PaLM-2 have made tremendous progress in solving fundamental math problems like GSM8K by achieving over 90% accuracy. However, their capabilities to solve more challenging math problems which require domain-specific knowledge (i.e. theorem) have yet to be investigated. In this paper, we introduce TheoremQA, the first theorem-driven question-answering dataset designed to evaluate AI models' capabilities to apply theorems to solve challenging science problems. TheoremQA is curated by domain experts containing 800 high-quality questions covering 350 theorems (e.g. Ta"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2305.12524","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-21T17:51:35Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c9b26605822aab58426a431df593154eb8db8cb69fec500c34a4f698936b92a4","abstract_canon_sha256":"7109df58195b71a72333019bef8cce2485d61bab0e61f8c6fa5cc64dd8fbb6e6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:20:50.950659Z","signature_b64":"9yJ9839/GasR/tDRcL5Bu8795mQjlCmsMOsWjLuJs4XKc1tZAPu+fuHSo7yAeSABPhH/MVT64xfpyZVtdHMPAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"12c925d8dda0cdb3711ef1bcd31ed077efd118a45f01956d83e6e59c586e3038","last_reissued_at":"2026-07-05T07:20:50.950164Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:20:50.950164Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TheoremQA: A Theorem-driven Question Answering dataset","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jianyu Xu, Max Ku, Ming Yin, Pan Lu, Tony Xia, Wenhu Chen, Xinyi Wang, Xueguang Ma, Yixin Wan","submitted_at":"2023-05-21T17:51:35Z","abstract_excerpt":"The recent LLMs like GPT-4 and PaLM-2 have made tremendous progress in solving fundamental math problems like GSM8K by achieving over 90% accuracy. However, their capabilities to solve more challenging math problems which require domain-specific knowledge (i.e. theorem) have yet to be investigated. In this paper, we introduce TheoremQA, the first theorem-driven question-answering dataset designed to evaluate AI models' capabilities to apply theorems to solve challenging science problems. TheoremQA is curated by domain experts containing 800 high-quality questions covering 350 theorems (e.g. Ta"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.12524","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2305.12524/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2305.12524","created_at":"2026-07-05T07:20:50.950222+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.12524v3","created_at":"2026-07-05T07:20:50.950222+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.12524","created_at":"2026-07-05T07:20:50.950222+00:00"},{"alias_kind":"pith_short_12","alias_value":"CLESLWG5UDG3","created_at":"2026-07-05T07:20:50.950222+00:00"},{"alias_kind":"pith_short_16","alias_value":"CLESLWG5UDG3G4I6","created_at":"2026-07-05T07:20:50.950222+00:00"},{"alias_kind":"pith_short_8","alias_value":"CLESLWG5","created_at":"2026-07-05T07:20:50.950222+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11182","citing_title":"EEVEE: Towards Test-time Prompt Learning in the Real World for Self-Improving Agents","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31455","citing_title":"DRIFT: Decoupled Rollouts and Importance-Weighted Fine-Tuning for Efficient Multi-Turn Optimization","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2309.05653","citing_title":"MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuning","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2308.14132","citing_title":"Detecting Language Model Attacks with Perplexity","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13330","citing_title":"FIND: Toward Multimodal Financial Reasoning and Question Answering for Indic Languages","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03893","citing_title":"FeynmanBench: Benchmarking Multimodal LLMs on Diagrammatic Physics Reasoning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08904","citing_title":"OPT-BENCH: Evaluating the Iterative Self-Optimization of LLM Agents in Large-Scale Search Spaces","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09012","citing_title":"Re$^2$Math: Benchmarking Theorem Retrieval in Research-Level Mathematics","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12610","citing_title":"Transforming External Knowledge into Triplets for Enhanced Retrieval in RAG of LLMs","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20659","citing_title":"GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective Reasoning","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7","json":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7.json","graph_json":"https://pith.science/api/pith-number/CLESLWG5UDG3G4I66G6NGHWQO7/graph.json","events_json":"https://pith.science/api/pith-number/CLESLWG5UDG3G4I66G6NGHWQO7/events.json","paper":"https://pith.science/paper/CLESLWG5"},"agent_actions":{"view_html":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7","download_json":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7.json","view_paper":"https://pith.science/paper/CLESLWG5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.12524&json=true","fetch_graph":"https://pith.science/api/pith-number/CLESLWG5UDG3G4I66G6NGHWQO7/graph.json","fetch_events":"https://pith.science/api/pith-number/CLESLWG5UDG3G4I66G6NGHWQO7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7/action/storage_attestation","attest_author":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7/action/author_attestation","sign_citation":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7/action/citation_signature","submit_replication":"https://pith.science/pith/CLESLWG5UDG3G4I66G6NGHWQO7/action/replication_record"}},"created_at":"2026-07-05T07:20:50.950222+00:00","updated_at":"2026-07-05T07:20:50.950222+00:00"}