{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JEUWW4EPV5DYPCDYBYUPR3PAIJ","short_pith_number":"pith:JEUWW4EP","schema_version":"1.0","canonical_sha256":"49296b708faf478788780e28f8ede04251eef4e05b7470f38e96a700c4558a84","source":{"kind":"arxiv","id":"2505.11827","version":2},"attestation_state":"computed","paper":{"title":"Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hao Liu, Jun Fang, Naiqiang Tan, Wei Li, Yansong Ning","submitted_at":"2025-05-17T04:26:39Z","abstract_excerpt":"Compressing long chain-of-thought (CoT) from large language models (LLMs) is an emerging strategy to improve the reasoning efficiency of LLMs. Despite its promising benefits, existing studies equally compress all thoughts within a long CoT, hindering more concise and effective reasoning. To this end, we first investigate the importance of different thoughts by examining their effectiveness and efficiency in contributing to reasoning through automatic long CoT chunking and Monte Carlo rollouts. Building upon the insights, we propose a theoretically bounded metric to jointly measure the effectiv"},"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":"2505.11827","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-17T04:26:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f0a5398be113fc893d00ac517423c2a14d7da0dd701eef7bc45357d640f16f40","abstract_canon_sha256":"9ac3c03f8c0bc3de9b8612e347e732f2b20a29e5cc5a0bb314f78a1db21b425f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:13.601182Z","signature_b64":"UcJEgHI/fTaqqWld5lY6mivpL+R3nSDeriqXGtp6RB4y0QctLqV1zjM678OW797C991hbZWL61WbfIS1VKipDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"49296b708faf478788780e28f8ede04251eef4e05b7470f38e96a700c4558a84","last_reissued_at":"2026-07-05T11:09:13.600656Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:13.600656Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Not All Thoughts are Generated Equal: Efficient LLM Reasoning via Multi-Turn Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hao Liu, Jun Fang, Naiqiang Tan, Wei Li, Yansong Ning","submitted_at":"2025-05-17T04:26:39Z","abstract_excerpt":"Compressing long chain-of-thought (CoT) from large language models (LLMs) is an emerging strategy to improve the reasoning efficiency of LLMs. Despite its promising benefits, existing studies equally compress all thoughts within a long CoT, hindering more concise and effective reasoning. To this end, we first investigate the importance of different thoughts by examining their effectiveness and efficiency in contributing to reasoning through automatic long CoT chunking and Monte Carlo rollouts. Building upon the insights, we propose a theoretically bounded metric to jointly measure the effectiv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11827","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/2505.11827/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":"2505.11827","created_at":"2026-07-05T11:09:13.600725+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.11827v2","created_at":"2026-07-05T11:09:13.600725+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11827","created_at":"2026-07-05T11:09:13.600725+00:00"},{"alias_kind":"pith_short_12","alias_value":"JEUWW4EPV5DY","created_at":"2026-07-05T11:09:13.600725+00:00"},{"alias_kind":"pith_short_16","alias_value":"JEUWW4EPV5DYPCDY","created_at":"2026-07-05T11:09:13.600725+00:00"},{"alias_kind":"pith_short_8","alias_value":"JEUWW4EP","created_at":"2026-07-05T11:09:13.600725+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.21260","citing_title":"On the Cost and Benefit of Chain of Thought: A Learning-Theoretic Perspective","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2503.16419","citing_title":"Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models","ref_index":134,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06165","citing_title":"Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost","ref_index":216,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ","json":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ.json","graph_json":"https://pith.science/api/pith-number/JEUWW4EPV5DYPCDYBYUPR3PAIJ/graph.json","events_json":"https://pith.science/api/pith-number/JEUWW4EPV5DYPCDYBYUPR3PAIJ/events.json","paper":"https://pith.science/paper/JEUWW4EP"},"agent_actions":{"view_html":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ","download_json":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ.json","view_paper":"https://pith.science/paper/JEUWW4EP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.11827&json=true","fetch_graph":"https://pith.science/api/pith-number/JEUWW4EPV5DYPCDYBYUPR3PAIJ/graph.json","fetch_events":"https://pith.science/api/pith-number/JEUWW4EPV5DYPCDYBYUPR3PAIJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ/action/storage_attestation","attest_author":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ/action/author_attestation","sign_citation":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ/action/citation_signature","submit_replication":"https://pith.science/pith/JEUWW4EPV5DYPCDYBYUPR3PAIJ/action/replication_record"}},"created_at":"2026-07-05T11:09:13.600725+00:00","updated_at":"2026-07-05T11:09:13.600725+00:00"}