{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:IZY5DLX5VIM5P4GOKDIFBWYCYH","short_pith_number":"pith:IZY5DLX5","schema_version":"1.0","canonical_sha256":"4671d1aefdaa19d7f0ce50d050db02c1feba1215d69257ebf8f14b6acdf0b4e4","source":{"kind":"arxiv","id":"2402.13709","version":2},"attestation_state":"computed","paper":{"title":"SaGE: Evaluating Moral Consistency in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Manas Gaur, Ponnurangam Kumaraguru, Priyanshul Govil, Sreeram Vennam, Vamshi Krishna Bonagiri","submitted_at":"2024-02-21T11:23:21Z","abstract_excerpt":"Despite recent advancements showcasing the impressive capabilities of Large Language Models (LLMs) in conversational systems, we show that even state-of-the-art LLMs are morally inconsistent in their generations, questioning their reliability (and trustworthiness in general). Prior works in LLM evaluation focus on developing ground-truth data to measure accuracy on specific tasks. However, for moral scenarios that often lack universally agreed-upon answers, consistency in model responses becomes crucial for their reliability. To address this issue, we propose an information-theoretic measure c"},"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":"2402.13709","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-21T11:23:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1542b6c977e04698516e3b015489ef08c904840a953e7fe7d4bb645882bb50f4","abstract_canon_sha256":"bc494e8144517e798847a4e32622f2aff8406145289b8f31f3d086b4590011af"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:50.520421Z","signature_b64":"9kQZKloFl2vXryNVxsg7587SdFYbpGF9lnPEUll/+YNb2QlDmu+GczbnmVlY6838c1Y50ebuGK3HX7b7tE0FAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4671d1aefdaa19d7f0ce50d050db02c1feba1215d69257ebf8f14b6acdf0b4e4","last_reissued_at":"2026-07-05T07:53:50.520002Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:50.520002Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SaGE: Evaluating Moral Consistency in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Manas Gaur, Ponnurangam Kumaraguru, Priyanshul Govil, Sreeram Vennam, Vamshi Krishna Bonagiri","submitted_at":"2024-02-21T11:23:21Z","abstract_excerpt":"Despite recent advancements showcasing the impressive capabilities of Large Language Models (LLMs) in conversational systems, we show that even state-of-the-art LLMs are morally inconsistent in their generations, questioning their reliability (and trustworthiness in general). Prior works in LLM evaluation focus on developing ground-truth data to measure accuracy on specific tasks. However, for moral scenarios that often lack universally agreed-upon answers, consistency in model responses becomes crucial for their reliability. To address this issue, we propose an information-theoretic measure c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.13709","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/2402.13709/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":"2402.13709","created_at":"2026-07-05T07:53:50.520056+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.13709v2","created_at":"2026-07-05T07:53:50.520056+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.13709","created_at":"2026-07-05T07:53:50.520056+00:00"},{"alias_kind":"pith_short_12","alias_value":"IZY5DLX5VIM5","created_at":"2026-07-05T07:53:50.520056+00:00"},{"alias_kind":"pith_short_16","alias_value":"IZY5DLX5VIM5P4GO","created_at":"2026-07-05T07:53:50.520056+00:00"},{"alias_kind":"pith_short_8","alias_value":"IZY5DLX5","created_at":"2026-07-05T07:53:50.520056+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.02972","citing_title":"A Scalable Approach to Evaluating Moral Sensitivity in LLMs","ref_index":75,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH","json":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH.json","graph_json":"https://pith.science/api/pith-number/IZY5DLX5VIM5P4GOKDIFBWYCYH/graph.json","events_json":"https://pith.science/api/pith-number/IZY5DLX5VIM5P4GOKDIFBWYCYH/events.json","paper":"https://pith.science/paper/IZY5DLX5"},"agent_actions":{"view_html":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH","download_json":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH.json","view_paper":"https://pith.science/paper/IZY5DLX5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.13709&json=true","fetch_graph":"https://pith.science/api/pith-number/IZY5DLX5VIM5P4GOKDIFBWYCYH/graph.json","fetch_events":"https://pith.science/api/pith-number/IZY5DLX5VIM5P4GOKDIFBWYCYH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH/action/storage_attestation","attest_author":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH/action/author_attestation","sign_citation":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH/action/citation_signature","submit_replication":"https://pith.science/pith/IZY5DLX5VIM5P4GOKDIFBWYCYH/action/replication_record"}},"created_at":"2026-07-05T07:53:50.520056+00:00","updated_at":"2026-07-05T07:53:50.520056+00:00"}