{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:2IDE5ZYXKXRZDVGGHXYYLLA2BI","short_pith_number":"pith:2IDE5ZYX","canonical_record":{"source":{"id":"2502.05196","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-28T23:39:02Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"c09ce78a6c76228fe3a012280529e94c2fc1d8293f7d0fadebc1ed1df2d21a02","abstract_canon_sha256":"fd6c7ff2a968095cb4f1f23658bedef4f5007102f41dd49b914857ff4b248027"},"schema_version":"1.0"},"canonical_sha256":"d2064ee71755e391d4c63df185ac1a0a35a1e9f1e8a5bbb5af7cff70a39609b3","source":{"kind":"arxiv","id":"2502.05196","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.05196","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"arxiv_version","alias_value":"2502.05196v1","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.05196","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_12","alias_value":"2IDE5ZYXKXRZ","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_16","alias_value":"2IDE5ZYXKXRZDVGG","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_8","alias_value":"2IDE5ZYX","created_at":"2026-07-05T10:11:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:2IDE5ZYXKXRZDVGGHXYYLLA2BI","target":"record","payload":{"canonical_record":{"source":{"id":"2502.05196","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-28T23:39:02Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"c09ce78a6c76228fe3a012280529e94c2fc1d8293f7d0fadebc1ed1df2d21a02","abstract_canon_sha256":"fd6c7ff2a968095cb4f1f23658bedef4f5007102f41dd49b914857ff4b248027"},"schema_version":"1.0"},"canonical_sha256":"d2064ee71755e391d4c63df185ac1a0a35a1e9f1e8a5bbb5af7cff70a39609b3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:11:33.934224Z","signature_b64":"8yM3018LpsV+2nHVNEqaGtEAVfMHFoO9353T9NqMrHT8KJyvm4Klg9RvpUXpbxpotbmhD4xNEbbO4so/Nu3SCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d2064ee71755e391d4c63df185ac1a0a35a1e9f1e8a5bbb5af7cff70a39609b3","last_reissued_at":"2026-07-05T10:11:33.933821Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:11:33.933821Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.05196","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:11:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h18tBWhzQZMMhgylCxhy1nOOGsFcGMbtirwaZXogbV7qzqna8Slo3z8uKLedlXSD2/0RbYDQq6MmJnkIEYCKDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T17:19:43.238581Z"},"content_sha256":"b82718dcb63921c3ae6f19bd1f246d5c50aec93a2f32c605725a591d1fda24fa","schema_version":"1.0","event_id":"sha256:b82718dcb63921c3ae6f19bd1f246d5c50aec93a2f32c605725a591d1fda24fa"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:2IDE5ZYXKXRZDVGGHXYYLLA2BI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLMs Provide Unstable Answers to Legal Questions","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CY"],"primary_cat":"cs.CL","authors_text":"Andrew Blair-Stanek, Benjamin Van Durme","submitted_at":"2025-01-28T23:39:02Z","abstract_excerpt":"An LLM is stable if it reaches the same conclusion when asked the identical question multiple times. We find leading LLMs like gpt-4o, claude-3.5, and gemini-1.5 are unstable when providing answers to hard legal questions, even when made as deterministic as possible by setting temperature to 0. We curate and release a novel dataset of 500 legal questions distilled from real cases, involving two parties, with facts, competing legal arguments, and the question of which party should prevail. When provided the exact same question, we observe that LLMs sometimes say one party should win, while othe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.05196","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/2502.05196/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:11:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BNQe7R7kWT8KqTRQlm1B3UJmo4uqm9HsskSySueUj5TEWYdni9RyY51KjP8mxe+p4W60cei/tEAUp4StCjcgAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T17:19:43.238955Z"},"content_sha256":"13ec91d202f22eda6cf784989b0c29d0e3ba0d712c356115d454756323419495","schema_version":"1.0","event_id":"sha256:13ec91d202f22eda6cf784989b0c29d0e3ba0d712c356115d454756323419495"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/bundle.json","state_url":"https://pith.science/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-06T17:19:43Z","links":{"resolver":"https://pith.science/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI","bundle":"https://pith.science/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/bundle.json","state":"https://pith.science/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2IDE5ZYXKXRZDVGGHXYYLLA2BI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:2IDE5ZYXKXRZDVGGHXYYLLA2BI","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"fd6c7ff2a968095cb4f1f23658bedef4f5007102f41dd49b914857ff4b248027","cross_cats_sorted":["cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-28T23:39:02Z","title_canon_sha256":"c09ce78a6c76228fe3a012280529e94c2fc1d8293f7d0fadebc1ed1df2d21a02"},"schema_version":"1.0","source":{"id":"2502.05196","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.05196","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"arxiv_version","alias_value":"2502.05196v1","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.05196","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_12","alias_value":"2IDE5ZYXKXRZ","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_16","alias_value":"2IDE5ZYXKXRZDVGG","created_at":"2026-07-05T10:11:33Z"},{"alias_kind":"pith_short_8","alias_value":"2IDE5ZYX","created_at":"2026-07-05T10:11:33Z"}],"graph_snapshots":[{"event_id":"sha256:13ec91d202f22eda6cf784989b0c29d0e3ba0d712c356115d454756323419495","target":"graph","created_at":"2026-07-05T10:11:33Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.05196/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"An LLM is stable if it reaches the same conclusion when asked the identical question multiple times. We find leading LLMs like gpt-4o, claude-3.5, and gemini-1.5 are unstable when providing answers to hard legal questions, even when made as deterministic as possible by setting temperature to 0. We curate and release a novel dataset of 500 legal questions distilled from real cases, involving two parties, with facts, competing legal arguments, and the question of which party should prevail. When provided the exact same question, we observe that LLMs sometimes say one party should win, while othe","authors_text":"Andrew Blair-Stanek, Benjamin Van Durme","cross_cats":["cs.CY"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-28T23:39:02Z","title":"LLMs Provide Unstable Answers to Legal Questions"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.05196","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:b82718dcb63921c3ae6f19bd1f246d5c50aec93a2f32c605725a591d1fda24fa","target":"record","created_at":"2026-07-05T10:11:33Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"fd6c7ff2a968095cb4f1f23658bedef4f5007102f41dd49b914857ff4b248027","cross_cats_sorted":["cs.CY"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-28T23:39:02Z","title_canon_sha256":"c09ce78a6c76228fe3a012280529e94c2fc1d8293f7d0fadebc1ed1df2d21a02"},"schema_version":"1.0","source":{"id":"2502.05196","kind":"arxiv","version":1}},"canonical_sha256":"d2064ee71755e391d4c63df185ac1a0a35a1e9f1e8a5bbb5af7cff70a39609b3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d2064ee71755e391d4c63df185ac1a0a35a1e9f1e8a5bbb5af7cff70a39609b3","first_computed_at":"2026-07-05T10:11:33.933821Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:11:33.933821Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8yM3018LpsV+2nHVNEqaGtEAVfMHFoO9353T9NqMrHT8KJyvm4Klg9RvpUXpbxpotbmhD4xNEbbO4so/Nu3SCw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:11:33.934224Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.05196","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b82718dcb63921c3ae6f19bd1f246d5c50aec93a2f32c605725a591d1fda24fa","sha256:13ec91d202f22eda6cf784989b0c29d0e3ba0d712c356115d454756323419495"],"state_sha256":"ca8c15d6ff38f2da8070db21d3ea4e7c586a341b737bc36540f993832a739097"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VofJki4i0qiuRaYJGJWDL26FuknaVNPlzt289jwc3aPPvuzR7bu2RJrVIlBZ+PhLl54cXr8j7Ov8k79R2bjSAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T17:19:43.240964Z","bundle_sha256":"675ffb3834e4c9b0215d62154f30c6362d1b826b2bf1f11d4ef1c5232d9d9528"}}