{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:EWU56QZOKNWZFMEPYF57EBP24I","short_pith_number":"pith:EWU56QZO","canonical_record":{"source":{"id":"2504.21202","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-29T22:16:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6c276c5dcd3efc11f2b312fac21543011ac9f126b9acb49de688cd098351919d","abstract_canon_sha256":"f879c682e88cfcca9787bc928f11e784f26ed2d3ab35aff05ce51ffa630d177e"},"schema_version":"1.0"},"canonical_sha256":"25a9df432e536d92b08fc17bf205fae20bca2d902b8d6f297b38afe150fdb4d1","source":{"kind":"arxiv","id":"2504.21202","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.21202","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"arxiv_version","alias_value":"2504.21202v1","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.21202","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_12","alias_value":"EWU56QZOKNWZ","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_16","alias_value":"EWU56QZOKNWZFMEP","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_8","alias_value":"EWU56QZO","created_at":"2026-07-05T10:56:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:EWU56QZOKNWZFMEPYF57EBP24I","target":"record","payload":{"canonical_record":{"source":{"id":"2504.21202","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-29T22:16:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6c276c5dcd3efc11f2b312fac21543011ac9f126b9acb49de688cd098351919d","abstract_canon_sha256":"f879c682e88cfcca9787bc928f11e784f26ed2d3ab35aff05ce51ffa630d177e"},"schema_version":"1.0"},"canonical_sha256":"25a9df432e536d92b08fc17bf205fae20bca2d902b8d6f297b38afe150fdb4d1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:56:19.244930Z","signature_b64":"jkJoUuvEuJXhfN+W7ddSwHBreQZCvyOlrMX6u7s5fhL3UFodKNaeqD1oD0fECmQuFijKshyuXUWpIs2JtrgrAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"25a9df432e536d92b08fc17bf205fae20bca2d902b8d6f297b38afe150fdb4d1","last_reissued_at":"2026-07-05T10:56:19.244393Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:56:19.244393Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.21202","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":"2026-07-05T10:56:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"msAB0lcdDJOJRVk4CT40p6a686Emz2FtTbSrzqNfcXf+yp/mj9QI68D2X6m09iIlTLcy/ReX8p+3CU0+YUFYDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T09:48:43.449645Z"},"content_sha256":"cfed5867e352d31347b19e94178e379d0c9794e49b9e9f7dbba774c3934b2296","schema_version":"1.0","event_id":"sha256:cfed5867e352d31347b19e94178e379d0c9794e49b9e9f7dbba774c3934b2296"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:EWU56QZOKNWZFMEPYF57EBP24I","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Automatic Legal Writing Evaluation of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ramon Pires, Rodrigo Nogueira, Roseval Malaquias Junior","submitted_at":"2025-04-29T22:16:39Z","abstract_excerpt":"Despite the recent advances in Large Language Models, benchmarks for evaluating legal writing remain scarce due to the inherent complexity of assessing open-ended responses in this domain. One of the key challenges in evaluating language models on domain-specific tasks is finding test datasets that are public, frequently updated, and contain comprehensive evaluation guidelines. The Brazilian Bar Examination meets these requirements. We introduce oab-bench, a benchmark comprising 105 questions across seven areas of law from recent editions of the exam. The benchmark includes comprehensive evalu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.21202","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/2504.21202/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":"2026-07-05T10:56:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PFlDoNeaVeiwqrffqoanpfZ4pe5aIm3RoDtODhIzE4b0B5Sg9C2iMXgz7CEs8wIxHFy9Sck4p28INRujU0CZCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T09:48:43.450158Z"},"content_sha256":"5caefcd4fcd1c44f99e411929efebc231bcee2aeb2c51bf212a92a5170d6f23d","schema_version":"1.0","event_id":"sha256:5caefcd4fcd1c44f99e411929efebc231bcee2aeb2c51bf212a92a5170d6f23d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EWU56QZOKNWZFMEPYF57EBP24I/bundle.json","state_url":"https://pith.science/pith/EWU56QZOKNWZFMEPYF57EBP24I/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EWU56QZOKNWZFMEPYF57EBP24I/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":"2026-08-10T09:48:43Z","links":{"resolver":"https://pith.science/pith/EWU56QZOKNWZFMEPYF57EBP24I","bundle":"https://pith.science/pith/EWU56QZOKNWZFMEPYF57EBP24I/bundle.json","state":"https://pith.science/pith/EWU56QZOKNWZFMEPYF57EBP24I/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EWU56QZOKNWZFMEPYF57EBP24I/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:EWU56QZOKNWZFMEPYF57EBP24I","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":"f879c682e88cfcca9787bc928f11e784f26ed2d3ab35aff05ce51ffa630d177e","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-29T22:16:39Z","title_canon_sha256":"6c276c5dcd3efc11f2b312fac21543011ac9f126b9acb49de688cd098351919d"},"schema_version":"1.0","source":{"id":"2504.21202","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.21202","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"arxiv_version","alias_value":"2504.21202v1","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.21202","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_12","alias_value":"EWU56QZOKNWZ","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_16","alias_value":"EWU56QZOKNWZFMEP","created_at":"2026-07-05T10:56:19Z"},{"alias_kind":"pith_short_8","alias_value":"EWU56QZO","created_at":"2026-07-05T10:56:19Z"}],"graph_snapshots":[{"event_id":"sha256:5caefcd4fcd1c44f99e411929efebc231bcee2aeb2c51bf212a92a5170d6f23d","target":"graph","created_at":"2026-07-05T10:56:19Z","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/2504.21202/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite the recent advances in Large Language Models, benchmarks for evaluating legal writing remain scarce due to the inherent complexity of assessing open-ended responses in this domain. One of the key challenges in evaluating language models on domain-specific tasks is finding test datasets that are public, frequently updated, and contain comprehensive evaluation guidelines. The Brazilian Bar Examination meets these requirements. We introduce oab-bench, a benchmark comprising 105 questions across seven areas of law from recent editions of the exam. The benchmark includes comprehensive evalu","authors_text":"Ramon Pires, Rodrigo Nogueira, Roseval Malaquias Junior","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-29T22:16:39Z","title":"Automatic Legal Writing Evaluation of LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.21202","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:cfed5867e352d31347b19e94178e379d0c9794e49b9e9f7dbba774c3934b2296","target":"record","created_at":"2026-07-05T10:56:19Z","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":"f879c682e88cfcca9787bc928f11e784f26ed2d3ab35aff05ce51ffa630d177e","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-29T22:16:39Z","title_canon_sha256":"6c276c5dcd3efc11f2b312fac21543011ac9f126b9acb49de688cd098351919d"},"schema_version":"1.0","source":{"id":"2504.21202","kind":"arxiv","version":1}},"canonical_sha256":"25a9df432e536d92b08fc17bf205fae20bca2d902b8d6f297b38afe150fdb4d1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"25a9df432e536d92b08fc17bf205fae20bca2d902b8d6f297b38afe150fdb4d1","first_computed_at":"2026-07-05T10:56:19.244393Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:56:19.244393Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jkJoUuvEuJXhfN+W7ddSwHBreQZCvyOlrMX6u7s5fhL3UFodKNaeqD1oD0fECmQuFijKshyuXUWpIs2JtrgrAw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:56:19.244930Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.21202","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cfed5867e352d31347b19e94178e379d0c9794e49b9e9f7dbba774c3934b2296","sha256:5caefcd4fcd1c44f99e411929efebc231bcee2aeb2c51bf212a92a5170d6f23d"],"state_sha256":"70e82532082d65fb0f1ced045e3dd020e075c31250350dc900c6e7993c6db9e7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FqD8oh538LJwXBlCA6s7G94PyDkTInwXSEDSIqnVIwUxJ3t0Bk9wBSJUEPjbsljcuUlWBEPBZLJ2o3rq+u5fCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T09:48:43.456039Z","bundle_sha256":"2b1b7dd7bdaecfd009aed4cb8c9b9599f296235e850dd5a03491a5655813680a"}}