{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QBAYXLYYU4XAXBO5ISF6F7XHOU","short_pith_number":"pith:QBAYXLYY","schema_version":"1.0","canonical_sha256":"80418baf18a72e0b85dd448be2fee7752ae0176199921878bb6357b1d2bd03c8","source":{"kind":"arxiv","id":"2505.23480","version":1},"attestation_state":"computed","paper":{"title":"Revisiting Overthinking in Long Chain-of-Thought from the Perspective of Self-Doubt","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dacheng Tao, Keqin Peng, Liang Ding, Meng Fang, Yuanxin Ouyang","submitted_at":"2025-05-29T14:30:02Z","abstract_excerpt":"Reasoning Large Language Models (RLLMs) have demonstrated impressive performance on complex tasks, largely due to the adoption of Long Chain-of-Thought (Long CoT) reasoning. However, they often exhibit overthinking -- performing unnecessary reasoning steps even after arriving at the correct answer. Prior work has largely focused on qualitative analyses of overthinking through sample-based observations of long CoTs. In contrast, we present a quantitative analysis of overthinking from the perspective of self-doubt, characterized by excessive token usage devoted to re-verifying already-correct an"},"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":"2505.23480","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-29T14:30:02Z","cross_cats_sorted":[],"title_canon_sha256":"895440d8699b271860fcc462e8d1c8636985f6093477f25fd09c765d3425f876","abstract_canon_sha256":"042c3b5224c9af242acf36d95d513ffe5d800b9b3d74ab589994b59216833612"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:02.678228Z","signature_b64":"YsbKBUECweGGaDLap08kZ81LMwh8K6etYZuNWk20cpvLCcUKyCpZc/c9NUsUVwk+9D9osCXIm2Qy8aqNPi8ABg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"80418baf18a72e0b85dd448be2fee7752ae0176199921878bb6357b1d2bd03c8","last_reissued_at":"2026-07-05T11:12:02.677560Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:02.677560Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Revisiting Overthinking in Long Chain-of-Thought from the Perspective of Self-Doubt","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dacheng Tao, Keqin Peng, Liang Ding, Meng Fang, Yuanxin Ouyang","submitted_at":"2025-05-29T14:30:02Z","abstract_excerpt":"Reasoning Large Language Models (RLLMs) have demonstrated impressive performance on complex tasks, largely due to the adoption of Long Chain-of-Thought (Long CoT) reasoning. However, they often exhibit overthinking -- performing unnecessary reasoning steps even after arriving at the correct answer. Prior work has largely focused on qualitative analyses of overthinking through sample-based observations of long CoTs. In contrast, we present a quantitative analysis of overthinking from the perspective of self-doubt, characterized by excessive token usage devoted to re-verifying already-correct an"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23480","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/2505.23480/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":"2505.23480","created_at":"2026-07-05T11:12:02.677637+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.23480v1","created_at":"2026-07-05T11:12:02.677637+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23480","created_at":"2026-07-05T11:12:02.677637+00:00"},{"alias_kind":"pith_short_12","alias_value":"QBAYXLYYU4XA","created_at":"2026-07-05T11:12:02.677637+00:00"},{"alias_kind":"pith_short_16","alias_value":"QBAYXLYYU4XAXBO5","created_at":"2026-07-05T11:12:02.677637+00:00"},{"alias_kind":"pith_short_8","alias_value":"QBAYXLYY","created_at":"2026-07-05T11:12:02.677637+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.00482","citing_title":"Know When to Stop: Segment-Level Credit Assignment for Reducing Overthinking","ref_index":74,"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":36,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17890","citing_title":"Dynamic Rollout Editing for Reducing Overthinking in RL-Trained Reasoning Models","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22211","citing_title":"CLORE: Content-Level Optimization for Reasoning Efficiency","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU","json":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU.json","graph_json":"https://pith.science/api/pith-number/QBAYXLYYU4XAXBO5ISF6F7XHOU/graph.json","events_json":"https://pith.science/api/pith-number/QBAYXLYYU4XAXBO5ISF6F7XHOU/events.json","paper":"https://pith.science/paper/QBAYXLYY"},"agent_actions":{"view_html":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU","download_json":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU.json","view_paper":"https://pith.science/paper/QBAYXLYY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.23480&json=true","fetch_graph":"https://pith.science/api/pith-number/QBAYXLYYU4XAXBO5ISF6F7XHOU/graph.json","fetch_events":"https://pith.science/api/pith-number/QBAYXLYYU4XAXBO5ISF6F7XHOU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU/action/storage_attestation","attest_author":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU/action/author_attestation","sign_citation":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU/action/citation_signature","submit_replication":"https://pith.science/pith/QBAYXLYYU4XAXBO5ISF6F7XHOU/action/replication_record"}},"created_at":"2026-07-05T11:12:02.677637+00:00","updated_at":"2026-07-05T11:12:02.677637+00:00"}