{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6RCXNNVGS4KKOCJAOLMZJTMK6N","short_pith_number":"pith:6RCXNNVG","schema_version":"1.0","canonical_sha256":"f44576b6a69714a7092072d994cd8af358e3fea715022425e00dd771ea1f1f58","source":{"kind":"arxiv","id":"2401.11467","version":2},"attestation_state":"computed","paper":{"title":"Over-Reasoning and Redundant Calculation of Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Cheng-Han Chiang, Hung-yi Lee","submitted_at":"2024-01-21T11:42:18Z","abstract_excerpt":"Large language models (LLMs) can solve problems step-by-step. While this chain-of-thought (CoT) reasoning boosts LLMs' performance, it is unclear if LLMs \\textit{know} when to use CoT and whether those CoT are always necessary to answer the question. This paper shows that LLMs tend to generate redundant calculations and reasoning on a manually constructed math QA dataset, GSM8K-Zero. GSM8K-Zero is constructed such that the questions can be answered without any calculations, but LLMs, including Llama-2 models and Claude-2, tend to generate lengthy and unnecessary calculations to answer the ques"},"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":"2401.11467","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-21T11:42:18Z","cross_cats_sorted":[],"title_canon_sha256":"d7c4d4d907b38ad9d200ce210dfd0d710e8d878e9f84362cb345bfc04298c4c4","abstract_canon_sha256":"f3fa81671dda088bbd5fbffad53b36d10dbecf77832a93b790637b222dd076ca"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:58:26.901197Z","signature_b64":"nJolbqAMzigbE42zZBGDL+V3McHEYZAzRR8EPGIs4rt7kgP4ww4P6jIHvagqWZjAZ0CxwmIbYUpPsirUws13DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f44576b6a69714a7092072d994cd8af358e3fea715022425e00dd771ea1f1f58","last_reissued_at":"2026-07-05T07:58:26.900535Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:58:26.900535Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Over-Reasoning and Redundant Calculation of Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Cheng-Han Chiang, Hung-yi Lee","submitted_at":"2024-01-21T11:42:18Z","abstract_excerpt":"Large language models (LLMs) can solve problems step-by-step. While this chain-of-thought (CoT) reasoning boosts LLMs' performance, it is unclear if LLMs \\textit{know} when to use CoT and whether those CoT are always necessary to answer the question. This paper shows that LLMs tend to generate redundant calculations and reasoning on a manually constructed math QA dataset, GSM8K-Zero. GSM8K-Zero is constructed such that the questions can be answered without any calculations, but LLMs, including Llama-2 models and Claude-2, tend to generate lengthy and unnecessary calculations to answer the ques"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.11467","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/2401.11467/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":"2401.11467","created_at":"2026-07-05T07:58:26.900701+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.11467v2","created_at":"2026-07-05T07:58:26.900701+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.11467","created_at":"2026-07-05T07:58:26.900701+00:00"},{"alias_kind":"pith_short_12","alias_value":"6RCXNNVGS4KK","created_at":"2026-07-05T07:58:26.900701+00:00"},{"alias_kind":"pith_short_16","alias_value":"6RCXNNVGS4KKOCJA","created_at":"2026-07-05T07:58:26.900701+00:00"},{"alias_kind":"pith_short_8","alias_value":"6RCXNNVG","created_at":"2026-07-05T07:58:26.900701+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29799","citing_title":"The CRISTAL Method: Neurosymbolic analysis from AI-synthesized world models","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N","json":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N.json","graph_json":"https://pith.science/api/pith-number/6RCXNNVGS4KKOCJAOLMZJTMK6N/graph.json","events_json":"https://pith.science/api/pith-number/6RCXNNVGS4KKOCJAOLMZJTMK6N/events.json","paper":"https://pith.science/paper/6RCXNNVG"},"agent_actions":{"view_html":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N","download_json":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N.json","view_paper":"https://pith.science/paper/6RCXNNVG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.11467&json=true","fetch_graph":"https://pith.science/api/pith-number/6RCXNNVGS4KKOCJAOLMZJTMK6N/graph.json","fetch_events":"https://pith.science/api/pith-number/6RCXNNVGS4KKOCJAOLMZJTMK6N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N/action/storage_attestation","attest_author":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N/action/author_attestation","sign_citation":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N/action/citation_signature","submit_replication":"https://pith.science/pith/6RCXNNVGS4KKOCJAOLMZJTMK6N/action/replication_record"}},"created_at":"2026-07-05T07:58:26.900701+00:00","updated_at":"2026-07-05T07:58:26.900701+00:00"}