{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DHNPYQ5G2DWUV3T2BAPT6HOBNY","short_pith_number":"pith:DHNPYQ5G","schema_version":"1.0","canonical_sha256":"19dafc43a6d0ed4aee7a081f3f1dc16e2690bce9e729cb67e4cf0fedc999d5d7","source":{"kind":"arxiv","id":"2504.20571","version":3},"attestation_state":"computed","paper":{"title":"Reinforcement Learning for Reasoning in Large Language Models with One Training Example","license":"http://creativecommons.org/licenses/by/4.0/","headline":"One training example via reinforcement learning lifts an LLM's math reasoning score from 36% to 74% on MATH500.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Baolin Peng, Hao Cheng, Jianfeng Gao, Kuan Wang, Liliang Ren, Liyuan Liu, Qing Yang, Shuohang Wang, Simon Shaolei Du, Weizhu Chen, Xuehai He, Yelong Shen, Yiping Wang, Zhiyuan Zeng","submitted_at":"2025-04-29T09:24:30Z","abstract_excerpt":"We show that reinforcement learning with verifiable reward using one training example (1-shot RLVR) is effective in incentivizing the math reasoning capabilities of large language models (LLMs). Applying RLVR to the base model Qwen2.5-Math-1.5B, we identify a single example that elevates model performance on MATH500 from 36.0% to 73.6% (8.6% improvement beyond format correction), and improves the average performance across six common mathematical reasoning benchmarks from 17.6% to 35.7% (7.0% non-format gain). This result matches the performance obtained using the 1.2k DeepScaleR subset (MATH5"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2504.20571","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-04-29T09:24:30Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"a3ee8a00abb8653ecd09efee5e6e3dff17022749f7ceb3f9a19d8760fc0ac677","abstract_canon_sha256":"6438bc3f9b0ad57edc0c56bf97ba63f2ed3cd3b4b78e640459a294c532011eea"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:50.357719Z","signature_b64":"CDcBAeWaHXOcNc6o8DYNgDNh+UFk5jeLsgAZXk00XQtDJf8k+rul2TPAweNEkphMes3LZWedgsqG0QSWrWZoAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19dafc43a6d0ed4aee7a081f3f1dc16e2690bce9e729cb67e4cf0fedc999d5d7","last_reissued_at":"2026-05-17T23:38:50.357211Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:50.357211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reinforcement Learning for Reasoning in Large Language Models with One Training Example","license":"http://creativecommons.org/licenses/by/4.0/","headline":"One training example via reinforcement learning lifts an LLM's math reasoning score from 36% to 74% on MATH500.","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Baolin Peng, Hao Cheng, Jianfeng Gao, Kuan Wang, Liliang Ren, Liyuan Liu, Qing Yang, Shuohang Wang, Simon Shaolei Du, Weizhu Chen, Xuehai He, Yelong Shen, Yiping Wang, Zhiyuan Zeng","submitted_at":"2025-04-29T09:24:30Z","abstract_excerpt":"We show that reinforcement learning with verifiable reward using one training example (1-shot RLVR) is effective in incentivizing the math reasoning capabilities of large language models (LLMs). Applying RLVR to the base model Qwen2.5-Math-1.5B, we identify a single example that elevates model performance on MATH500 from 36.0% to 73.6% (8.6% improvement beyond format correction), and improves the average performance across six common mathematical reasoning benchmarks from 17.6% to 35.7% (7.0% non-format gain). This result matches the performance obtained using the 1.2k DeepScaleR subset (MATH5"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Reinforcement learning with verifiable reward using one training example (1-shot RLVR) is effective in incentivizing the math reasoning capabilities of large language models (LLMs). Applying RLVR to the base model Qwen2.5-Math-1.5B, we identify a single example that elevates model performance on MATH500 from 36.0% to 73.6%.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the single chosen example is not specially selected in a way that inflates generalization, and that the observed gains arise from the RL policy gradient rather than incidental effects of the training setup or prompt format.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"One training example via RLVR boosts LLM math reasoning from 17.6% to 35.7% average across six benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"One training example via reinforcement learning lifts an LLM's math reasoning score from 36% to 74% on MATH500.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"36735173058404e2fa8e7aad18c2de165a2691a58c91a23b42e5aca735632a56"},"source":{"id":"2504.20571","kind":"arxiv","version":3},"verdict":{"id":"c7b86817-aa42-4484-a016-6876f0248f25","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T19:47:48.868083Z","strongest_claim":"Reinforcement learning with verifiable reward using one training example (1-shot RLVR) is effective in incentivizing the math reasoning capabilities of large language models (LLMs). Applying RLVR to the base model Qwen2.5-Math-1.5B, we identify a single example that elevates model performance on MATH500 from 36.0% to 73.6%.","one_line_summary":"One training example via RLVR boosts LLM math reasoning from 17.6% to 35.7% average across six benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the single chosen example is not specially selected in a way that inflates generalization, and that the observed gains arise from the RL policy gradient rather than incidental effects of the training setup or prompt format.","pith_extraction_headline":"One training example via reinforcement learning lifts an LLM's math reasoning score from 36% to 74% on MATH500."},"references":{"count":69,"sample":[{"doi":"","year":2024,"title":"Learning to reason with llms","work_id":"170ab9a7-f02d-4838-a217-516dcecfaaf1","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning","work_id":"e6b75ad5-2877-4168-97c8-710407094d20","ref_index":2,"cited_arxiv_id":"2501.12948","is_internal_anchor":true},{"doi":"","year":2025,"title":"Kimi k1.5: Scaling Reinforcement Learning with LLMs","work_id":"bff96ab1-bd6a-4585-be23-74fdb51969c7","ref_index":3,"cited_arxiv_id":"2501.12599","is_internal_anchor":true},{"doi":"","year":2024,"title":"On designing effective rl reward at training time for llm reasoning","work_id":"5c46cc3b-db37-4655-a0ff-f28ab6c32a1f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Tulu 3: Pushing Frontiers in Open Language Model Post-Training","work_id":"28c9dbea-056a-48c2-8000-85f809827e45","ref_index":5,"cited_arxiv_id":"2411.15124","is_internal_anchor":true}],"resolved_work":69,"snapshot_sha256":"57a5e9920dcc9893c568b300a373c6224244244cb7150ad9389f9bf9eeae1bad","internal_anchors":29},"formal_canon":{"evidence_count":1,"snapshot_sha256":"dbbdd7d28668dbf699e7d875b7e1415989d324785463f554cf12a2052ab56764"},"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":"2504.20571","created_at":"2026-05-17T23:38:50.357293+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.20571v3","created_at":"2026-05-17T23:38:50.357293+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.20571","created_at":"2026-05-17T23:38:50.357293+00:00"},{"alias_kind":"pith_short_12","alias_value":"DHNPYQ5G2DWU","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"DHNPYQ5G2DWUV3T2","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"DHNPYQ5G","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":39,"internal_anchor_count":39,"sample":[{"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":165,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22642","citing_title":"Spreadsheet-RL: Advancing Large Language Model Agents on Realistic Spreadsheet Tasks via Reinforcement Learning","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2506.12119","citing_title":"Mixture-of-Experts Can Surpass Dense LLMs Under Strictly Equal Resource","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2602.01970","citing_title":"Small Generalizable Prompt Predictive Models Can Steer Efficient RL Post-Training of Large Reasoning Models","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20256","citing_title":"FBOS-RL: Feedback-Driven Bi-Objective Synergistic Reinforcement Learning","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2604.28005","citing_title":"Kernelized Advantage Estimation: From Nonparametric Statistics to LLM Reasoning","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16874","citing_title":"Reasoning Can Be Restored by Correcting a Few Decision Tokens","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2507.02833","citing_title":"Generalizing Verifiable Instruction Following","ref_index":29,"is_internal_anchor":true},{"citing_arxiv_id":"2507.12549","citing_title":"The Serial Scaling Hypothesis","ref_index":123,"is_internal_anchor":true},{"citing_arxiv_id":"2508.00222","citing_title":"RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2505.15134","citing_title":"The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoning","ref_index":88,"is_internal_anchor":true},{"citing_arxiv_id":"2509.25454","citing_title":"DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2510.10649","citing_title":"Unlocking Exploration in RLVR: Uncertainty-aware Advantage Shaping for Deeper Reasoning","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2511.00066","citing_title":"Sharpness-Guided Group Relative Policy Optimization via Probability Shaping","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2512.13399","citing_title":"Differentiable Evolutionary Reinforcement Learning","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2601.20829","citing_title":"Training Reasoning Models on Saturated Problems via Failure-Prefix Conditioning","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2601.21464","citing_title":"Conversation for Non-verifiable Learning: Self-Evolving LLMs through Meta-Evaluation","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2602.03452","citing_title":"Beyond Variance: Prompt-Efficient RLVR via Rare-Event Amplification and Bidirectional Pairing","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2604.02341","citing_title":"LLM Reasoning with Process Rewards for Outcome-Guided Steps","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2601.16175","citing_title":"Learning to Discover at Test Time","ref_index":79,"is_internal_anchor":true},{"citing_arxiv_id":"2506.21734","citing_title":"Hierarchical Reasoning Model","ref_index":99,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11538","citing_title":"Taming Extreme Tokens: Covariance-Aware GRPO with Gaussian-Kernel Advantage Reweighting","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2506.01939","citing_title":"Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2604.28020","citing_title":"Cost-Aware Learning","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2604.28005","citing_title":"Kernelized Advantage Estimation: From Nonparametric Statistics to LLM Reasoning","ref_index":24,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY","json":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY.json","graph_json":"https://pith.science/api/pith-number/DHNPYQ5G2DWUV3T2BAPT6HOBNY/graph.json","events_json":"https://pith.science/api/pith-number/DHNPYQ5G2DWUV3T2BAPT6HOBNY/events.json","paper":"https://pith.science/paper/DHNPYQ5G"},"agent_actions":{"view_html":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY","download_json":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY.json","view_paper":"https://pith.science/paper/DHNPYQ5G","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.20571&json=true","fetch_graph":"https://pith.science/api/pith-number/DHNPYQ5G2DWUV3T2BAPT6HOBNY/graph.json","fetch_events":"https://pith.science/api/pith-number/DHNPYQ5G2DWUV3T2BAPT6HOBNY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY/action/storage_attestation","attest_author":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY/action/author_attestation","sign_citation":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY/action/citation_signature","submit_replication":"https://pith.science/pith/DHNPYQ5G2DWUV3T2BAPT6HOBNY/action/replication_record"}},"created_at":"2026-05-17T23:38:50.357293+00:00","updated_at":"2026-05-17T23:38:50.357293+00:00"}