{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:YEBWTL32ROCKEEYS5KFUPILYIP","short_pith_number":"pith:YEBWTL32","canonical_record":{"source":{"id":"2505.11733","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T22:34:36Z","cross_cats_sorted":[],"title_canon_sha256":"4794b45d6ef1562acb2c37bb2eb3bbe809339ff1a70851e6da0f3694e89a96ae","abstract_canon_sha256":"db278be9f29dc52cdb873bacfaf3ad1934064453a222786723c0f5f1830ecf11"},"schema_version":"1.0"},"canonical_sha256":"c10369af7a8b84a21312ea8b47a17843e706cafc056c00998c62af19d82314d1","source":{"kind":"arxiv","id":"2505.11733","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.11733","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"arxiv_version","alias_value":"2505.11733v2","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11733","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_12","alias_value":"YEBWTL32ROCK","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_16","alias_value":"YEBWTL32ROCKEEYS","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_8","alias_value":"YEBWTL32","created_at":"2026-07-05T11:06:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:YEBWTL32ROCKEEYS5KFUPILYIP","target":"record","payload":{"canonical_record":{"source":{"id":"2505.11733","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T22:34:36Z","cross_cats_sorted":[],"title_canon_sha256":"4794b45d6ef1562acb2c37bb2eb3bbe809339ff1a70851e6da0f3694e89a96ae","abstract_canon_sha256":"db278be9f29dc52cdb873bacfaf3ad1934064453a222786723c0f5f1830ecf11"},"schema_version":"1.0"},"canonical_sha256":"c10369af7a8b84a21312ea8b47a17843e706cafc056c00998c62af19d82314d1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:13.616423Z","signature_b64":"M8D6X65QbOAJYrWJ/iBnR3hmrcf4aniEobG/dIIAoPe6HhIZQPhtH42wKv5GQ9vm+VC6Ga7tqDjUjVt6OL+4DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c10369af7a8b84a21312ea8b47a17843e706cafc056c00998c62af19d82314d1","last_reissued_at":"2026-07-05T11:06:13.615892Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:13.615892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.11733","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:06:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uBof4qv/Y14Q//3qYSOP+LpoSrujICq8SUtJhruS9l/gUI3rjRTohzI07jy8rdiIUuS+LQGnujjKgEnJ2E7OAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T20:15:12.319223Z"},"content_sha256":"604bbffce7064289c4988cb3b643f9cd5e1bb6a93a14c87c30bf69faf6de1638","schema_version":"1.0","event_id":"sha256:604bbffce7064289c4988cb3b643f9cd5e1bb6a93a14c87c30bf69faf6de1638"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:YEBWTL32ROCKEEYS5KFUPILYIP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MedCaseReasoning: Evaluating and learning diagnostic reasoning from clinical case reports","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alejandro Lozano, Angela Zhang, Arvind Suresh, Eric Wu, Jacqueline J. Tao, James Zou, Kevin Wei, Kevin Wu, Min Woo Sun, Rahul Thapa","submitted_at":"2025-05-16T22:34:36Z","abstract_excerpt":"Doctors and patients alike increasingly use Large Language Models (LLMs) to diagnose clinical cases. However, unlike domains such as math or coding, where correctness can be objectively defined by the final answer, medical diagnosis requires both the outcome and the reasoning process to be accurate. Currently, widely used medical benchmarks like MedQA and MMLU assess only accuracy in the final answer, overlooking the quality and faithfulness of the clinical reasoning process. To address this limitation, we introduce MedCaseReasoning, the first open-access dataset for evaluating LLMs on their a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11733","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/2505.11733/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":"2026-07-05T11:06:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"z8vGn2x6PANaWNSMdSgR6rZkpvcwz3WCsEpK+Cv+MEyK4KKUpj2etA0Ka3HVKb83mr6nr8aGDyAc0BHj0v45AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T20:15:12.319844Z"},"content_sha256":"1a4f39cd52f41a9e60899afb77aef12d085dae1cf42bd016e4377a55c6d766bb","schema_version":"1.0","event_id":"sha256:1a4f39cd52f41a9e60899afb77aef12d085dae1cf42bd016e4377a55c6d766bb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YEBWTL32ROCKEEYS5KFUPILYIP/bundle.json","state_url":"https://pith.science/pith/YEBWTL32ROCKEEYS5KFUPILYIP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YEBWTL32ROCKEEYS5KFUPILYIP/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":"2026-08-04T20:15:12Z","links":{"resolver":"https://pith.science/pith/YEBWTL32ROCKEEYS5KFUPILYIP","bundle":"https://pith.science/pith/YEBWTL32ROCKEEYS5KFUPILYIP/bundle.json","state":"https://pith.science/pith/YEBWTL32ROCKEEYS5KFUPILYIP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YEBWTL32ROCKEEYS5KFUPILYIP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:YEBWTL32ROCKEEYS5KFUPILYIP","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":"db278be9f29dc52cdb873bacfaf3ad1934064453a222786723c0f5f1830ecf11","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T22:34:36Z","title_canon_sha256":"4794b45d6ef1562acb2c37bb2eb3bbe809339ff1a70851e6da0f3694e89a96ae"},"schema_version":"1.0","source":{"id":"2505.11733","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.11733","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"arxiv_version","alias_value":"2505.11733v2","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.11733","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_12","alias_value":"YEBWTL32ROCK","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_16","alias_value":"YEBWTL32ROCKEEYS","created_at":"2026-07-05T11:06:13Z"},{"alias_kind":"pith_short_8","alias_value":"YEBWTL32","created_at":"2026-07-05T11:06:13Z"}],"graph_snapshots":[{"event_id":"sha256:1a4f39cd52f41a9e60899afb77aef12d085dae1cf42bd016e4377a55c6d766bb","target":"graph","created_at":"2026-07-05T11:06:13Z","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/2505.11733/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Doctors and patients alike increasingly use Large Language Models (LLMs) to diagnose clinical cases. However, unlike domains such as math or coding, where correctness can be objectively defined by the final answer, medical diagnosis requires both the outcome and the reasoning process to be accurate. Currently, widely used medical benchmarks like MedQA and MMLU assess only accuracy in the final answer, overlooking the quality and faithfulness of the clinical reasoning process. To address this limitation, we introduce MedCaseReasoning, the first open-access dataset for evaluating LLMs on their a","authors_text":"Alejandro Lozano, Angela Zhang, Arvind Suresh, Eric Wu, Jacqueline J. Tao, James Zou, Kevin Wei, Kevin Wu, Min Woo Sun, Rahul Thapa","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T22:34:36Z","title":"MedCaseReasoning: Evaluating and learning diagnostic reasoning from clinical case reports"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.11733","kind":"arxiv","version":2},"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":"sha256:604bbffce7064289c4988cb3b643f9cd5e1bb6a93a14c87c30bf69faf6de1638","target":"record","created_at":"2026-07-05T11:06:13Z","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":"db278be9f29dc52cdb873bacfaf3ad1934064453a222786723c0f5f1830ecf11","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-16T22:34:36Z","title_canon_sha256":"4794b45d6ef1562acb2c37bb2eb3bbe809339ff1a70851e6da0f3694e89a96ae"},"schema_version":"1.0","source":{"id":"2505.11733","kind":"arxiv","version":2}},"canonical_sha256":"c10369af7a8b84a21312ea8b47a17843e706cafc056c00998c62af19d82314d1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c10369af7a8b84a21312ea8b47a17843e706cafc056c00998c62af19d82314d1","first_computed_at":"2026-07-05T11:06:13.615892Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:06:13.615892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"M8D6X65QbOAJYrWJ/iBnR3hmrcf4aniEobG/dIIAoPe6HhIZQPhtH42wKv5GQ9vm+VC6Ga7tqDjUjVt6OL+4DA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:06:13.616423Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.11733","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:604bbffce7064289c4988cb3b643f9cd5e1bb6a93a14c87c30bf69faf6de1638","sha256:1a4f39cd52f41a9e60899afb77aef12d085dae1cf42bd016e4377a55c6d766bb"],"state_sha256":"28c8a49b4d6f8d6e76f879464006a6bb5c00e33633e9fcc14a5b188751fafa8b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YbAiaenlQUaN+ppRKI6kxIKlf7Ma8N6RuZegoPAC/zWVzqiv+CiwZ8q0YtVBZUQwuapPKP3MjbDUTrg70RCcBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T20:15:12.353532Z","bundle_sha256":"655a33df16ef6c9eff7719e92e602f1df03a69f1986bd27645342cfc8ec552b2"}}