{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KWJVDNEIIVNRYHEO6H35DYOC7W","short_pith_number":"pith:KWJVDNEI","schema_version":"1.0","canonical_sha256":"559351b488455b1c1c8ef1f7d1e1c2fdb4cb7651cd166cc380d29f235b2cef7c","source":{"kind":"arxiv","id":"2506.08343","version":2},"attestation_state":"computed","paper":{"title":"Wait, We Don't Need to \"Wait\"! Removing Thinking Tokens Improves Reasoning Efficiency","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenlong Wang, Dongping Chen, Ranjay Krishna, Tianyi Zhou, Yuanning Feng, Zhaoyang Chu","submitted_at":"2025-06-10T01:54:04Z","abstract_excerpt":"Recent advances in large reasoning models have enabled complex, step-by-step reasoning but often introduce significant overthinking, resulting in verbose and redundant outputs that hinder efficiency. In this study, we examine whether explicit self-reflection, signaled by tokens such as \"Wait\" and \"Hmm\", is necessary for advanced reasoning. We propose NoWait, a simple yet effective approach that disables explicit self-reflection by suppressing these tokens during inference. Extensive experiments on ten benchmarks across textual, visual, and video reasoning tasks show that NoWait reduces chain-o"},"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":"2506.08343","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-10T01:54:04Z","cross_cats_sorted":[],"title_canon_sha256":"9cbc9d75862f3345dbba263492bd1820404ea36dd53e1c6a4c1f7d6c9a11d8b1","abstract_canon_sha256":"cd82a8f0fc959e39fae0b0d7a93a883f647b19d86a93e2cfb8a21b0890dfee02"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:29.584546Z","signature_b64":"YhtmjoiPn59fcIHe4YdMGxbeXx3wEiuA+mwxR4P7ZzlpnkkP7z82/Oc/4x+OLm2spBowr2SimQCMwE4e3jTrDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"559351b488455b1c1c8ef1f7d1e1c2fdb4cb7651cd166cc380d29f235b2cef7c","last_reissued_at":"2026-07-05T11:23:29.584041Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:29.584041Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Wait, We Don't Need to \"Wait\"! Removing Thinking Tokens Improves Reasoning Efficiency","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Chenlong Wang, Dongping Chen, Ranjay Krishna, Tianyi Zhou, Yuanning Feng, Zhaoyang Chu","submitted_at":"2025-06-10T01:54:04Z","abstract_excerpt":"Recent advances in large reasoning models have enabled complex, step-by-step reasoning but often introduce significant overthinking, resulting in verbose and redundant outputs that hinder efficiency. In this study, we examine whether explicit self-reflection, signaled by tokens such as \"Wait\" and \"Hmm\", is necessary for advanced reasoning. We propose NoWait, a simple yet effective approach that disables explicit self-reflection by suppressing these tokens during inference. Extensive experiments on ten benchmarks across textual, visual, and video reasoning tasks show that NoWait reduces chain-o"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.08343","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/2506.08343/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":"2506.08343","created_at":"2026-07-05T11:23:29.584095+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.08343v2","created_at":"2026-07-05T11:23:29.584095+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.08343","created_at":"2026-07-05T11:23:29.584095+00:00"},{"alias_kind":"pith_short_12","alias_value":"KWJVDNEIIVNR","created_at":"2026-07-05T11:23:29.584095+00:00"},{"alias_kind":"pith_short_16","alias_value":"KWJVDNEIIVNRYHEO","created_at":"2026-07-05T11:23:29.584095+00:00"},{"alias_kind":"pith_short_8","alias_value":"KWJVDNEI","created_at":"2026-07-05T11:23:29.584095+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22716","citing_title":"Beyond Penalizing Mistakes: Stabilizing Efficiency Training in Large Reasoning Models via Adaptive Correct-Only Rewards","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07108","citing_title":"DyCon: Dynamic Reasoning Control via Evolving Difficulty Modeling","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28305","citing_title":"Revisiting Anthropomorphic Reflection Markers in Large Language Model Reasoning","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2602.08324","citing_title":"Towards Efficient Large Language Reasoning Models via Extreme-Ratio Chain-of-Thought Compression","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2508.13654","citing_title":"Input-Time Scaling: Adding Noise and Irrelevance into Less-Is-More Drastically Improves Reasoning Performance and Efficiency","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2603.05433","citing_title":"CRISP: Compressed Reasoning via Iterative Self-Policy Distillation","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21764","citing_title":"Thinking with Reasoning Skills: Fewer Tokens, More Accuracy","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06787","citing_title":"When Is Thinking Enough? Early Exit via Sufficiency Assessment for Efficient Reasoning","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05655","citing_title":"LLM Reasoning as Trajectories: Step-Specific Representation Geometry and Correctness Signals","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05643","citing_title":"Graph-Based Chain-of-Thought Pruning for Reducing Redundant Reflections in Reasoning LLMs","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W","json":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W.json","graph_json":"https://pith.science/api/pith-number/KWJVDNEIIVNRYHEO6H35DYOC7W/graph.json","events_json":"https://pith.science/api/pith-number/KWJVDNEIIVNRYHEO6H35DYOC7W/events.json","paper":"https://pith.science/paper/KWJVDNEI"},"agent_actions":{"view_html":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W","download_json":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W.json","view_paper":"https://pith.science/paper/KWJVDNEI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.08343&json=true","fetch_graph":"https://pith.science/api/pith-number/KWJVDNEIIVNRYHEO6H35DYOC7W/graph.json","fetch_events":"https://pith.science/api/pith-number/KWJVDNEIIVNRYHEO6H35DYOC7W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W/action/storage_attestation","attest_author":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W/action/author_attestation","sign_citation":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W/action/citation_signature","submit_replication":"https://pith.science/pith/KWJVDNEIIVNRYHEO6H35DYOC7W/action/replication_record"}},"created_at":"2026-07-05T11:23:29.584095+00:00","updated_at":"2026-07-05T11:23:29.584095+00:00"}