{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:4RY3WEJCESNQ5PQR3BPZ4BM5FB","short_pith_number":"pith:4RY3WEJC","schema_version":"1.0","canonical_sha256":"e471bb1122249b0ebe11d85f9e059d285aee682fa03306313a40a8ac589b9ab8","source":{"kind":"arxiv","id":"2203.14465","version":2},"attestation_state":"computed","paper":{"title":"STaR: Bootstrapping Reasoning With Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Eric Zelikman, Jesse Mu, Noah D. Goodman, Yuhuai Wu","submitted_at":"2022-03-28T03:12:15Z","abstract_excerpt":"Generating step-by-step \"chain-of-thought\" rationales improves language model performance on complex reasoning tasks like mathematics or commonsense question-answering. However, inducing language model rationale generation currently requires either constructing massive rationale datasets or sacrificing accuracy by using only few-shot inference. We propose a technique to iteratively leverage a small number of rationale examples and a large dataset without rationales, to bootstrap the ability to perform successively more complex reasoning. This technique, the \"Self-Taught Reasoner\" (STaR), relie"},"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":"2203.14465","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-28T03:12:15Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"c3b5f7cc873770b6d7c24ec9e5fbe464f5f9cdafe7a2d7c07750df45618a3379","abstract_canon_sha256":"8e7333fe783ef254c3bd71158b52c2dcd96d73192b00d6f5cbbf3dbdf621f6d6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:24:59.083491Z","signature_b64":"C/b/Hal90G1b2NKlsFKAY31d/Xs8fjt+YyXz0QXQoD6retEMJX6wU1qfhWX1lCTQ20t5SbhGbH3J5baypKUOBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e471bb1122249b0ebe11d85f9e059d285aee682fa03306313a40a8ac589b9ab8","last_reissued_at":"2026-07-05T04:24:59.082982Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:24:59.082982Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"STaR: Bootstrapping Reasoning With Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Eric Zelikman, Jesse Mu, Noah D. Goodman, Yuhuai Wu","submitted_at":"2022-03-28T03:12:15Z","abstract_excerpt":"Generating step-by-step \"chain-of-thought\" rationales improves language model performance on complex reasoning tasks like mathematics or commonsense question-answering. However, inducing language model rationale generation currently requires either constructing massive rationale datasets or sacrificing accuracy by using only few-shot inference. We propose a technique to iteratively leverage a small number of rationale examples and a large dataset without rationales, to bootstrap the ability to perform successively more complex reasoning. This technique, the \"Self-Taught Reasoner\" (STaR), relie"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.14465","kind":"arxiv","version":2},"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/2203.14465/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":"2203.14465","created_at":"2026-07-05T04:24:59.083045+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.14465v2","created_at":"2026-07-05T04:24:59.083045+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.14465","created_at":"2026-07-05T04:24:59.083045+00:00"},{"alias_kind":"pith_short_12","alias_value":"4RY3WEJCESNQ","created_at":"2026-07-05T04:24:59.083045+00:00"},{"alias_kind":"pith_short_16","alias_value":"4RY3WEJCESNQ5PQR","created_at":"2026-07-05T04:24:59.083045+00:00"},{"alias_kind":"pith_short_8","alias_value":"4RY3WEJC","created_at":"2026-07-05T04:24:59.083045+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":39,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.07646","citing_title":"RL Post-Training Builds Compositional Reasoning Strategies","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2607.07663","citing_title":"Recursive Self-Improvement in AI: From Bounded Self-Refinement to Autonomous Research Loops","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2606.24937","citing_title":"The Hitchhiker's Guide to Agentic AI: From Foundations to Systems","ref_index":237,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22992","citing_title":"Predicate Importance Estimation and Decoupled Rationale-Score Distillation for Entity Alignment","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21884","citing_title":"A Verifiable Search Is Not a Learnable Chain-of-Thought","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21121","citing_title":"Answer Engineering: Local Trajectory Editing for Protocol-Constrained Decision Making in Large Language Models","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20295","citing_title":"Token-Operations-Oriented Inference Optimization Techniques for Large Models","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2606.18216","citing_title":"Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02460","citing_title":"Neuron-Aware Data Selection for Annotation-Free LLM Self-Distillation","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07856","citing_title":"Teacher-Free Self-Training Amplifies but Does Not Compound: A Pass@$K$ Crossover on a Free-Verifier Domain","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05253","citing_title":"Alpha-RTL: Test-Time Training for RTL Hardware Optimization","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03503","citing_title":"ThoughtFold: Folding Reasoning Chains via Introspective Preference Learning","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03938","citing_title":"q0: Primitives for Hyper-Epoch Pretraining","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00172","citing_title":"CAST: Non-Privileged Clipped Asymmetric Self-Teaching with Advantage Flipping for GRPO","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30775","citing_title":"A Single Rewrite Suffices: Empirical Lessons from Production Skill Description Optimization","ref_index":79,"is_internal_anchor":false},{"citing_arxiv_id":"2606.32017","citing_title":"TRIAGE: Role-Typed Credit Assignment for Agentic Reinforcement Learning","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28070","citing_title":"Bridging the Detection-to-Abstention Gap in Reasoning Models under Insufficient Information","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30914","citing_title":"Automating Formal Verification with Reinforcement Learning and Recursive Inference","ref_index":112,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00726","citing_title":"Latent Reward Steering: An Adaptive Inference-Time Framework that Implicitly Promotes Cognitive Behaviors in Reasoning LLMs","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02812","citing_title":"Traj-Evolve: A Self-Evolving Multi-Agent System for Patient Trajectory Modeling in Lung Cancer Early Detection","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07006","citing_title":"RASFT: Rollout-Adaptive Supervised Fine-Tuning for Reasoning","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2211.14275","citing_title":"Solving math word problems with process- and outcome-based feedback","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2211.01910","citing_title":"Large Language Models Are Human-Level Prompt Engineers","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2504.01990","citing_title":"Advances and Challenges in Foundation Agents: From Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems","ref_index":88,"is_internal_anchor":false},{"citing_arxiv_id":"2505.23678","citing_title":"Grounded Reinforcement Learning for Visual Reasoning","ref_index":86,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB","json":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB.json","graph_json":"https://pith.science/api/pith-number/4RY3WEJCESNQ5PQR3BPZ4BM5FB/graph.json","events_json":"https://pith.science/api/pith-number/4RY3WEJCESNQ5PQR3BPZ4BM5FB/events.json","paper":"https://pith.science/paper/4RY3WEJC"},"agent_actions":{"view_html":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB","download_json":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB.json","view_paper":"https://pith.science/paper/4RY3WEJC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.14465&json=true","fetch_graph":"https://pith.science/api/pith-number/4RY3WEJCESNQ5PQR3BPZ4BM5FB/graph.json","fetch_events":"https://pith.science/api/pith-number/4RY3WEJCESNQ5PQR3BPZ4BM5FB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB/action/storage_attestation","attest_author":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB/action/author_attestation","sign_citation":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB/action/citation_signature","submit_replication":"https://pith.science/pith/4RY3WEJCESNQ5PQR3BPZ4BM5FB/action/replication_record"}},"created_at":"2026-07-05T04:24:59.083045+00:00","updated_at":"2026-07-05T04:24:59.083045+00:00"}