{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6JRCQZHEFSTQ2YKGH7WKSBYGHW","short_pith_number":"pith:6JRCQZHE","schema_version":"1.0","canonical_sha256":"f2622864e42ca70d61463feca907063db8e9640d5e5d87e1a4d8190d2a332739","source":{"kind":"arxiv","id":"2402.03620","version":1},"attestation_state":"computed","paper":{"title":"Self-Discover: Large Language Models Self-Compose Reasoning Structures","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Denny Zhou, Ed H. Chi, Heng-Tze Cheng, Huaixiu Steven Zheng, Jay Pujara, Pei Zhou, Quoc V. Le, Swaroop Mishra, Xiang Ren, Xinyun Chen","submitted_at":"2024-02-06T01:13:53Z","abstract_excerpt":"We introduce SELF-DISCOVER, a general framework for LLMs to self-discover the task-intrinsic reasoning structures to tackle complex reasoning problems that are challenging for typical prompting methods. Core to the framework is a self-discovery process where LLMs select multiple atomic reasoning modules such as critical thinking and step-by-step thinking, and compose them into an explicit reasoning structure for LLMs to follow during decoding. SELF-DISCOVER substantially improves GPT-4 and PaLM 2's performance on challenging reasoning benchmarks such as BigBench-Hard, grounded agent reasoning,"},"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":"2402.03620","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2024-02-06T01:13:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"c7da8372263ad5ca18e033cdb455563f2d96e030c7c4ffbb03066a776cccd994","abstract_canon_sha256":"372f2c0db58e890dbffe6749c5b3e61dcb629c074c0772d0eff1fe370bb84c93"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:41:55.432580Z","signature_b64":"4D21mcNrPMO5wMwQg85UocALwcjrByioQv6/xkOQe/e3C15tnoEobr3ZF2Fjd+umSyw5t/jsg0N00zIdE7zWAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f2622864e42ca70d61463feca907063db8e9640d5e5d87e1a4d8190d2a332739","last_reissued_at":"2026-07-05T07:41:55.432109Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:41:55.432109Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Discover: Large Language Models Self-Compose Reasoning Structures","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Denny Zhou, Ed H. Chi, Heng-Tze Cheng, Huaixiu Steven Zheng, Jay Pujara, Pei Zhou, Quoc V. Le, Swaroop Mishra, Xiang Ren, Xinyun Chen","submitted_at":"2024-02-06T01:13:53Z","abstract_excerpt":"We introduce SELF-DISCOVER, a general framework for LLMs to self-discover the task-intrinsic reasoning structures to tackle complex reasoning problems that are challenging for typical prompting methods. Core to the framework is a self-discovery process where LLMs select multiple atomic reasoning modules such as critical thinking and step-by-step thinking, and compose them into an explicit reasoning structure for LLMs to follow during decoding. SELF-DISCOVER substantially improves GPT-4 and PaLM 2's performance on challenging reasoning benchmarks such as BigBench-Hard, grounded agent reasoning,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.03620","kind":"arxiv","version":1},"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/2402.03620/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":"2402.03620","created_at":"2026-07-05T07:41:55.432163+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.03620v1","created_at":"2026-07-05T07:41:55.432163+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.03620","created_at":"2026-07-05T07:41:55.432163+00:00"},{"alias_kind":"pith_short_12","alias_value":"6JRCQZHEFSTQ","created_at":"2026-07-05T07:41:55.432163+00:00"},{"alias_kind":"pith_short_16","alias_value":"6JRCQZHEFSTQ2YKG","created_at":"2026-07-05T07:41:55.432163+00:00"},{"alias_kind":"pith_short_8","alias_value":"6JRCQZHE","created_at":"2026-07-05T07:41:55.432163+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22610","citing_title":"PaperClaw: Harnessing Agents for Autonomous Research and Human-in-the-Loop Refinement","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02390","citing_title":"DecompRL: Solving Harder Problems by Learning Modular Code Generation","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00038","citing_title":"Stop Hand-Holding Your Coding Agent: Engineering the Loops that Replace Step-by-Step Prompting","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08704","citing_title":"AgentPSO: Evolving Agent Reasoning Skill via Multi-agent Particle Swarm Optimization","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2411.15594","citing_title":"A Survey on LLM-as-a-Judge","ref_index":227,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17187","citing_title":"PluRule: A Benchmark for Moderating Pluralistic Communities on Social Media","ref_index":199,"is_internal_anchor":false},{"citing_arxiv_id":"2408.08435","citing_title":"Automated Design of Agentic Systems","ref_index":238,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08486","citing_title":"Teachers' Perceived Benefits and Risks of AI Across Fifty-Five Countries: An Audit of LLM Alignment and Steerability","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08704","citing_title":"AgentPSO: Evolving Agent Reasoning Skill via Multi-agent Particle Swarm Optimization","ref_index":53,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW","json":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW.json","graph_json":"https://pith.science/api/pith-number/6JRCQZHEFSTQ2YKGH7WKSBYGHW/graph.json","events_json":"https://pith.science/api/pith-number/6JRCQZHEFSTQ2YKGH7WKSBYGHW/events.json","paper":"https://pith.science/paper/6JRCQZHE"},"agent_actions":{"view_html":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW","download_json":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW.json","view_paper":"https://pith.science/paper/6JRCQZHE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.03620&json=true","fetch_graph":"https://pith.science/api/pith-number/6JRCQZHEFSTQ2YKGH7WKSBYGHW/graph.json","fetch_events":"https://pith.science/api/pith-number/6JRCQZHEFSTQ2YKGH7WKSBYGHW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW/action/storage_attestation","attest_author":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW/action/author_attestation","sign_citation":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW/action/citation_signature","submit_replication":"https://pith.science/pith/6JRCQZHEFSTQ2YKGH7WKSBYGHW/action/replication_record"}},"created_at":"2026-07-05T07:41:55.432163+00:00","updated_at":"2026-07-05T07:41:55.432163+00:00"}