{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:6MQ7O56VW2M23M7F5LB32I5LDA","short_pith_number":"pith:6MQ7O56V","schema_version":"1.0","canonical_sha256":"f321f777d5b699adb3e5eac3bd23ab183e7c8f7d2a971222e689dd05a47cbc6f","source":{"kind":"arxiv","id":"2505.17746","version":2},"attestation_state":"computed","paper":{"title":"Fast Quiet-STaR: Thinking Without Thought Tokens","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guiguang Ding, Hui Chen, Wei Huang, Xin Ye, Yizhe Xiong, Zhijie Deng, Zijia Lin","submitted_at":"2025-05-23T11:14:12Z","abstract_excerpt":"Large Language Models (LLMs) have achieved impressive performance across a range of natural language processing tasks. However, recent advances demonstrate that further gains particularly in complex reasoning tasks require more than merely scaling up model sizes or training data. One promising direction is to enable models to think during the reasoning process. Recently, Quiet STaR significantly improves reasoning by generating token-level thought traces, but incurs substantial inference overhead. In this work, we propose Fast Quiet STaR, a more efficient reasoning framework that preserves the"},"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":"2505.17746","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-23T11:14:12Z","cross_cats_sorted":[],"title_canon_sha256":"4253aacdd2aa91e4233947378f043c6abc0762ee8237dc4b0c5cec3ac0db52a9","abstract_canon_sha256":"0538c01e6c34e55c787c5b8e406f785b56ee274e419ddef7ca8847f03af49f90"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:06:12.997233Z","signature_b64":"p1ZT1xmZEcmtHLCagq87JC7pEMR8bKPoBFgCkDgXirNrKcT2Ei22onLhFtHeExLVmjaVJ5Mh1jzb48s4/vpZAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f321f777d5b699adb3e5eac3bd23ab183e7c8f7d2a971222e689dd05a47cbc6f","last_reissued_at":"2026-07-05T12:06:12.996696Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:06:12.996696Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fast Quiet-STaR: Thinking Without Thought Tokens","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guiguang Ding, Hui Chen, Wei Huang, Xin Ye, Yizhe Xiong, Zhijie Deng, Zijia Lin","submitted_at":"2025-05-23T11:14:12Z","abstract_excerpt":"Large Language Models (LLMs) have achieved impressive performance across a range of natural language processing tasks. However, recent advances demonstrate that further gains particularly in complex reasoning tasks require more than merely scaling up model sizes or training data. One promising direction is to enable models to think during the reasoning process. Recently, Quiet STaR significantly improves reasoning by generating token-level thought traces, but incurs substantial inference overhead. In this work, we propose Fast Quiet STaR, a more efficient reasoning framework that preserves the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17746","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/2505.17746/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":"2505.17746","created_at":"2026-07-05T12:06:12.996761+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17746v2","created_at":"2026-07-05T12:06:12.996761+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17746","created_at":"2026-07-05T12:06:12.996761+00:00"},{"alias_kind":"pith_short_12","alias_value":"6MQ7O56VW2M2","created_at":"2026-07-05T12:06:12.996761+00:00"},{"alias_kind":"pith_short_16","alias_value":"6MQ7O56VW2M23M7F","created_at":"2026-07-05T12:06:12.996761+00:00"},{"alias_kind":"pith_short_8","alias_value":"6MQ7O56V","created_at":"2026-07-05T12:06:12.996761+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.18464","citing_title":"PERL: Parameter Efficient Reasoning in CLIP Latent Space","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA","json":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA.json","graph_json":"https://pith.science/api/pith-number/6MQ7O56VW2M23M7F5LB32I5LDA/graph.json","events_json":"https://pith.science/api/pith-number/6MQ7O56VW2M23M7F5LB32I5LDA/events.json","paper":"https://pith.science/paper/6MQ7O56V"},"agent_actions":{"view_html":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA","download_json":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA.json","view_paper":"https://pith.science/paper/6MQ7O56V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17746&json=true","fetch_graph":"https://pith.science/api/pith-number/6MQ7O56VW2M23M7F5LB32I5LDA/graph.json","fetch_events":"https://pith.science/api/pith-number/6MQ7O56VW2M23M7F5LB32I5LDA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA/action/storage_attestation","attest_author":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA/action/author_attestation","sign_citation":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA/action/citation_signature","submit_replication":"https://pith.science/pith/6MQ7O56VW2M23M7F5LB32I5LDA/action/replication_record"}},"created_at":"2026-07-05T12:06:12.996761+00:00","updated_at":"2026-07-05T12:06:12.996761+00:00"}