{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:EVC2OBZX4O3PGC23DABRETOYK6","short_pith_number":"pith:EVC2OBZX","schema_version":"1.0","canonical_sha256":"2545a70737e3b6f30b5b1803124dd857a6f89e19971508da20badeb8d457947c","source":{"kind":"arxiv","id":"2410.23856","version":1},"attestation_state":"computed","paper":{"title":"Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bo Han, Jianing Zhu, Rong Tao, Yiwen Luo, Zengmao Wang, Zhanke Zhou","submitted_at":"2024-10-31T12:07:44Z","abstract_excerpt":"This paper investigates an under-explored challenge in large language models (LLMs): chain-of-thought prompting with noisy rationales, which include irrelevant or inaccurate reasoning thoughts within examples used for in-context learning. We construct NoRa dataset that is tailored to evaluate the robustness of reasoning in the presence of noisy rationales. Our findings on NoRa dataset reveal a prevalent vulnerability to such noise among current LLMs, with existing robust methods like self-correction and self-consistency showing limited efficacy. Notably, compared to prompting with clean ration"},"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":"2410.23856","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-31T12:07:44Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"071344cb71833db907e534a0d448ec0c05c8ac14da9312bd817ccbcc98693daa","abstract_canon_sha256":"66333d6721c834ceb1c310cc0133588f8a48bd1e9d35d1d727d7b934a1e56adc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:29:06.137263Z","signature_b64":"XYqu/C+IhOisYICzCjaK9I+mFsL5BvYDAyyh25hILJ7LnzwVn78ZaYBE38uzNAFraoEndhVEf/wHvo0lAwM3DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2545a70737e3b6f30b5b1803124dd857a6f89e19971508da20badeb8d457947c","last_reissued_at":"2026-07-05T09:29:06.136757Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:29:06.136757Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Bo Han, Jianing Zhu, Rong Tao, Yiwen Luo, Zengmao Wang, Zhanke Zhou","submitted_at":"2024-10-31T12:07:44Z","abstract_excerpt":"This paper investigates an under-explored challenge in large language models (LLMs): chain-of-thought prompting with noisy rationales, which include irrelevant or inaccurate reasoning thoughts within examples used for in-context learning. We construct NoRa dataset that is tailored to evaluate the robustness of reasoning in the presence of noisy rationales. Our findings on NoRa dataset reveal a prevalent vulnerability to such noise among current LLMs, with existing robust methods like self-correction and self-consistency showing limited efficacy. Notably, compared to prompting with clean ration"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.23856","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/2410.23856/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":"2410.23856","created_at":"2026-07-05T09:29:06.136820+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.23856v1","created_at":"2026-07-05T09:29:06.136820+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.23856","created_at":"2026-07-05T09:29:06.136820+00:00"},{"alias_kind":"pith_short_12","alias_value":"EVC2OBZX4O3P","created_at":"2026-07-05T09:29:06.136820+00:00"},{"alias_kind":"pith_short_16","alias_value":"EVC2OBZX4O3PGC23","created_at":"2026-07-05T09:29:06.136820+00:00"},{"alias_kind":"pith_short_8","alias_value":"EVC2OBZX","created_at":"2026-07-05T09:29:06.136820+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.21642","citing_title":"Ablate-to-Validate: Are Vision-Language Models Really Using Continuous Thought Tokens?","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6","json":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6.json","graph_json":"https://pith.science/api/pith-number/EVC2OBZX4O3PGC23DABRETOYK6/graph.json","events_json":"https://pith.science/api/pith-number/EVC2OBZX4O3PGC23DABRETOYK6/events.json","paper":"https://pith.science/paper/EVC2OBZX"},"agent_actions":{"view_html":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6","download_json":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6.json","view_paper":"https://pith.science/paper/EVC2OBZX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.23856&json=true","fetch_graph":"https://pith.science/api/pith-number/EVC2OBZX4O3PGC23DABRETOYK6/graph.json","fetch_events":"https://pith.science/api/pith-number/EVC2OBZX4O3PGC23DABRETOYK6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6/action/storage_attestation","attest_author":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6/action/author_attestation","sign_citation":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6/action/citation_signature","submit_replication":"https://pith.science/pith/EVC2OBZX4O3PGC23DABRETOYK6/action/replication_record"}},"created_at":"2026-07-05T09:29:06.136820+00:00","updated_at":"2026-07-05T09:29:06.136820+00:00"}