{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MDR67XGRPNW3GVPDTGKK4KZEIP","short_pith_number":"pith:MDR67XGR","schema_version":"1.0","canonical_sha256":"60e3efdcd17b6db355e39994ae2b2443c2530c35dedbd5143df0d3929cb23147","source":{"kind":"arxiv","id":"2408.06195","version":1},"attestation_state":"computed","paper":{"title":"Mutual Reasoning Makes Smaller LLMs Stronger Problem-Solvers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fan Yang, Jiahang Xu, Li Lyna Zhang, Mao Yang, Mingyuan Ma, Zhenting Qi","submitted_at":"2024-08-12T14:42:13Z","abstract_excerpt":"This paper introduces rStar, a self-play mutual reasoning approach that significantly improves reasoning capabilities of small language models (SLMs) without fine-tuning or superior models. rStar decouples reasoning into a self-play mutual generation-discrimination process. First, a target SLM augments the Monte Carlo Tree Search (MCTS) with a rich set of human-like reasoning actions to construct higher quality reasoning trajectories. Next, another SLM, with capabilities similar to the target SLM, acts as a discriminator to verify each trajectory generated by the target SLM. The mutually agree"},"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":"2408.06195","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-08-12T14:42:13Z","cross_cats_sorted":[],"title_canon_sha256":"d7e07845f7229d3926f1ec350ed2d5426e87824c73e05f3faf5fdbe307ed0738","abstract_canon_sha256":"3d8187aac37ac75d461f80d7f85b29e9d1e0f294bd1ab9b51fec26388416159a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:54:37.675175Z","signature_b64":"A5YT/Xf2ODX/HuHi8hzJIOuLCC4aYaikxd/X1Cxs4qBPojQ4Q8eG6KcXn5mlEPSwsyAqzK+exVF/dOnAOnRjCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"60e3efdcd17b6db355e39994ae2b2443c2530c35dedbd5143df0d3929cb23147","last_reissued_at":"2026-07-05T08:54:37.674716Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:54:37.674716Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mutual Reasoning Makes Smaller LLMs Stronger Problem-Solvers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fan Yang, Jiahang Xu, Li Lyna Zhang, Mao Yang, Mingyuan Ma, Zhenting Qi","submitted_at":"2024-08-12T14:42:13Z","abstract_excerpt":"This paper introduces rStar, a self-play mutual reasoning approach that significantly improves reasoning capabilities of small language models (SLMs) without fine-tuning or superior models. rStar decouples reasoning into a self-play mutual generation-discrimination process. First, a target SLM augments the Monte Carlo Tree Search (MCTS) with a rich set of human-like reasoning actions to construct higher quality reasoning trajectories. Next, another SLM, with capabilities similar to the target SLM, acts as a discriminator to verify each trajectory generated by the target SLM. The mutually agree"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.06195","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/2408.06195/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":"2408.06195","created_at":"2026-07-05T08:54:37.674773+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.06195v1","created_at":"2026-07-05T08:54:37.674773+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.06195","created_at":"2026-07-05T08:54:37.674773+00:00"},{"alias_kind":"pith_short_12","alias_value":"MDR67XGRPNW3","created_at":"2026-07-05T08:54:37.674773+00:00"},{"alias_kind":"pith_short_16","alias_value":"MDR67XGRPNW3GVPD","created_at":"2026-07-05T08:54:37.674773+00:00"},{"alias_kind":"pith_short_8","alias_value":"MDR67XGR","created_at":"2026-07-05T08:54:37.674773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":22,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06974","citing_title":"MILES: Modular Instruction Memory with Learnable Selection for Self-Improving LLM Reasoning","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2606.24937","citing_title":"The Hitchhiker's Guide to Agentic AI: From Foundations to Systems","ref_index":254,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22385","citing_title":"MetaPS: Adaptive Programmatic Strategy Selection for Market Agents","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08068","citing_title":"DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination","ref_index":185,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01612","citing_title":"Self-Improving Small Object Grounding in LVLMs","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28566","citing_title":"Tree of Thoughts as a Classical Heuristic Search Problem: Formal Foundations and Design Patterns","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07006","citing_title":"RASFT: Rollout-Adaptive Supervised Fine-Tuning for Reasoning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2502.00955","citing_title":"Efficient Multi-Agent System Training with Data Influence-Oriented Tree Search","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2505.15858","citing_title":"Search-Based Multi-Trajectory Refinement for Safe C-to-Rust Translation with Large Language Models","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16302","citing_title":"Reducing Credit Assignment Variance via Counterfactual Reasoning Paths","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17295","citing_title":"DISA: Offline Importance Sampling for Distribution-Matching LLM-RL","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2509.23322","citing_title":"Mitigating Visual Context Degradation in Large Multimodal Models: A Training-Free Decoupled Agentic Framework","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2509.25454","citing_title":"DeepSearch: Overcome the Bottleneck of Reinforcement Learning with Verifiable Rewards via Monte Carlo Tree Search","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2510.18245","citing_title":"Scaling Laws Meet Model Architecture: Toward Inference-Efficient LLMs","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2512.14044","citing_title":"OmniDrive-R1: Reinforcement-driven Interleaved Multi-modal Chain-of-Thought for Trustworthy Vision-Language Autonomous Driving","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10195","citing_title":"Breaking the Reward Barrier: Accelerating Tree-of-Thought Reasoning via Speculative Exploration","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2502.17419","citing_title":"From System 1 to System 2: A Survey of Reasoning Large Language Models","ref_index":231,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10195","citing_title":"Breaking the Reward Barrier: Accelerating Tree-of-Thought Reasoning via Speculative Exploration","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09492","citing_title":"APCD: Adaptive Path-Contrastive Decoding for Reliable Large Language Model Generation","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.10228","citing_title":"SVSR: A Self-Verification and Self-Rectification Paradigm for Multimodal Reasoning","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04066","citing_title":"Adapt to Thrive! Adaptive Power-Mean Policy Optimization for Improved LLM Reasoning","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04065","citing_title":"Free Energy-Driven Reinforcement Learning with Adaptive Advantage Shaping for Unsupervised Reasoning in LLMs","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP","json":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP.json","graph_json":"https://pith.science/api/pith-number/MDR67XGRPNW3GVPDTGKK4KZEIP/graph.json","events_json":"https://pith.science/api/pith-number/MDR67XGRPNW3GVPDTGKK4KZEIP/events.json","paper":"https://pith.science/paper/MDR67XGR"},"agent_actions":{"view_html":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP","download_json":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP.json","view_paper":"https://pith.science/paper/MDR67XGR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.06195&json=true","fetch_graph":"https://pith.science/api/pith-number/MDR67XGRPNW3GVPDTGKK4KZEIP/graph.json","fetch_events":"https://pith.science/api/pith-number/MDR67XGRPNW3GVPDTGKK4KZEIP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP/action/storage_attestation","attest_author":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP/action/author_attestation","sign_citation":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP/action/citation_signature","submit_replication":"https://pith.science/pith/MDR67XGRPNW3GVPDTGKK4KZEIP/action/replication_record"}},"created_at":"2026-07-05T08:54:37.674773+00:00","updated_at":"2026-07-05T08:54:37.674773+00:00"}