{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3FNAFTDAWUKIIWOFVP34GZT6JR","short_pith_number":"pith:3FNAFTDA","schema_version":"1.0","canonical_sha256":"d95a02cc60b5148459c5abf7c3667e4c5bafb6e4d4cd5829bfa1e1dd2c99d53b","source":{"kind":"arxiv","id":"2503.02670","version":1},"attestation_state":"computed","paper":{"title":"Multidimensional Consistency Improves Reasoning in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Huiyuan Lai, Malvina Nissim, Xiao Zhang","submitted_at":"2025-03-04T14:41:05Z","abstract_excerpt":"While Large language models (LLMs) have proved able to address some complex reasoning tasks, we also know that they are highly sensitive to input variation, which can lead to different solution paths and final answers. Answer consistency across input variations can thus be taken as a sign of stronger confidence. Leveraging this insight, we introduce a framework, {\\em Multidimensional Reasoning Consistency} where, focusing on math problems, models are systematically pushed to diversify solution paths towards a final answer, thereby testing them for answer consistency across multiple input varia"},"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":"2503.02670","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-03-04T14:41:05Z","cross_cats_sorted":[],"title_canon_sha256":"c3dbd18db682b6bef065799aa8b8f15b9cb81b891946935fcd0d70e5bc6711aa","abstract_canon_sha256":"ebf50655560b0f5f63aa669aa5f34ce40c028dc7940e0babaedab1c8daca1c96"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:24:11.197839Z","signature_b64":"Wv2GHG/BC4vWA1deiC+2/CPCW8sake0GIklmO3l8vGelVCcaG7WBU7SWK1zagReJ5/1HQFv7PiTt31WZWVWcBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d95a02cc60b5148459c5abf7c3667e4c5bafb6e4d4cd5829bfa1e1dd2c99d53b","last_reissued_at":"2026-07-05T10:24:11.195236Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:24:11.195236Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multidimensional Consistency Improves Reasoning in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Huiyuan Lai, Malvina Nissim, Xiao Zhang","submitted_at":"2025-03-04T14:41:05Z","abstract_excerpt":"While Large language models (LLMs) have proved able to address some complex reasoning tasks, we also know that they are highly sensitive to input variation, which can lead to different solution paths and final answers. Answer consistency across input variations can thus be taken as a sign of stronger confidence. Leveraging this insight, we introduce a framework, {\\em Multidimensional Reasoning Consistency} where, focusing on math problems, models are systematically pushed to diversify solution paths towards a final answer, thereby testing them for answer consistency across multiple input varia"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.02670","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/2503.02670/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":"2503.02670","created_at":"2026-07-05T10:24:11.195290+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.02670v1","created_at":"2026-07-05T10:24:11.195290+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.02670","created_at":"2026-07-05T10:24:11.195290+00:00"},{"alias_kind":"pith_short_12","alias_value":"3FNAFTDAWUKI","created_at":"2026-07-05T10:24:11.195290+00:00"},{"alias_kind":"pith_short_16","alias_value":"3FNAFTDAWUKIIWOF","created_at":"2026-07-05T10:24:11.195290+00:00"},{"alias_kind":"pith_short_8","alias_value":"3FNAFTDA","created_at":"2026-07-05T10:24:11.195290+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.26644","citing_title":"When to Vote, When to Rewrite: Disagreement-Guided Strategy Routing for Test-Time Scaling","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20090","citing_title":"Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR","json":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR.json","graph_json":"https://pith.science/api/pith-number/3FNAFTDAWUKIIWOFVP34GZT6JR/graph.json","events_json":"https://pith.science/api/pith-number/3FNAFTDAWUKIIWOFVP34GZT6JR/events.json","paper":"https://pith.science/paper/3FNAFTDA"},"agent_actions":{"view_html":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR","download_json":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR.json","view_paper":"https://pith.science/paper/3FNAFTDA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.02670&json=true","fetch_graph":"https://pith.science/api/pith-number/3FNAFTDAWUKIIWOFVP34GZT6JR/graph.json","fetch_events":"https://pith.science/api/pith-number/3FNAFTDAWUKIIWOFVP34GZT6JR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR/action/storage_attestation","attest_author":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR/action/author_attestation","sign_citation":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR/action/citation_signature","submit_replication":"https://pith.science/pith/3FNAFTDAWUKIIWOFVP34GZT6JR/action/replication_record"}},"created_at":"2026-07-05T10:24:11.195290+00:00","updated_at":"2026-07-05T10:24:11.195290+00:00"}