{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:RRV4GFQO3BSNG7R7ZASS5CJPUX","short_pith_number":"pith:RRV4GFQO","canonical_record":{"source":{"id":"2405.07960","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-05-13T17:38:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"329298b668b7d8094011bd393da2ad3e586fdf5a58433485b44e89dc083faef2","abstract_canon_sha256":"6dab0129b3cd64aa7f6b361c04cad632a91ccee45f9bb199d24ecc97c73d5a18"},"schema_version":"1.0"},"canonical_sha256":"8c6bc3160ed864d37e3fc8252e892fa5fefc98956feda7b04dfc1e388a6f7a0e","source":{"kind":"arxiv","id":"2405.07960","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.07960","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"arxiv_version","alias_value":"2405.07960v5","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.07960","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_12","alias_value":"RRV4GFQO3BSN","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_16","alias_value":"RRV4GFQO3BSNG7R7","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_8","alias_value":"RRV4GFQO","created_at":"2026-05-21T13:47:14Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:RRV4GFQO3BSNG7R7ZASS5CJPUX","target":"record","payload":{"canonical_record":{"source":{"id":"2405.07960","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-05-13T17:38:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"329298b668b7d8094011bd393da2ad3e586fdf5a58433485b44e89dc083faef2","abstract_canon_sha256":"6dab0129b3cd64aa7f6b361c04cad632a91ccee45f9bb199d24ecc97c73d5a18"},"schema_version":"1.0"},"canonical_sha256":"8c6bc3160ed864d37e3fc8252e892fa5fefc98956feda7b04dfc1e388a6f7a0e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T13:47:14.140102Z","signature_b64":"rQf3bBRBTZsz/4Vr4rxTme1qf+PMR+V9DMPuguzUiFFhoYMqzvhhK5bbEGxVpNBjlpsLFCJG2z1GO9FDLmHSBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8c6bc3160ed864d37e3fc8252e892fa5fefc98956feda7b04dfc1e388a6f7a0e","last_reissued_at":"2026-05-21T13:47:14.138213Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T13:47:14.138213Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.07960","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-21T13:47:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"F+ssmJxleoWldtudSL8Jyulr69kXIZbnJtFnydw1La4xEO9WZ8tUueXddQ0jWu0bBxCzjezWPlXDnuuVNHfzBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T03:43:22.331984Z"},"content_sha256":"e4d12ce6135c05cfdffb1d770b40a88214ec2b860dad4c822a2e673cc70cd468","schema_version":"1.0","event_id":"sha256:e4d12ce6135c05cfdffb1d770b40a88214ec2b860dad4c822a2e673cc70cd468"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:RRV4GFQO3BSNG7R7ZASS5CJPUX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environments","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.HC","authors_text":"Carl Harris, Eduardo Reis, Jeffrey Jopling, Michael Moor, Rojin Ziaei, Samuel Schmidgall","submitted_at":"2024-05-13T17:38:53Z","abstract_excerpt":"Evaluating large language models (LLM) in clinical scenarios is crucial to assessing their potential clinical utility. Existing benchmarks rely heavily on static question-answering, which does not accurately depict the complex, sequential nature of clinical decision-making. Here, we introduce AgentClinic, a multimodal agent benchmark for evaluating LLMs in simulated clinical environments that include patient interactions, multimodal data collection under incomplete information, and the usage of various tools, resulting in an in-depth evaluation across nine medical specialties and seven languag"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.07960","kind":"arxiv","version":5},"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/2405.07960/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":"2026-05-21T13:47:14Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ynCQTd0mr4atpk4C437Xhm90THhEXlUVxly+7I6bTIaGV9FHwNQPMq43TxvvGgT3xz0AIkn/QOWRf4pftYZBDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T03:43:22.332475Z"},"content_sha256":"efede858b0b90b7831d28abaa802b1a3b243960b08e4a228db9f79fdb96d6215","schema_version":"1.0","event_id":"sha256:efede858b0b90b7831d28abaa802b1a3b243960b08e4a228db9f79fdb96d6215"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/bundle.json","state_url":"https://pith.science/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/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":"2026-05-25T03:43:22Z","links":{"resolver":"https://pith.science/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX","bundle":"https://pith.science/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/bundle.json","state":"https://pith.science/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RRV4GFQO3BSNG7R7ZASS5CJPUX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:RRV4GFQO3BSNG7R7ZASS5CJPUX","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":"6dab0129b3cd64aa7f6b361c04cad632a91ccee45f9bb199d24ecc97c73d5a18","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-05-13T17:38:53Z","title_canon_sha256":"329298b668b7d8094011bd393da2ad3e586fdf5a58433485b44e89dc083faef2"},"schema_version":"1.0","source":{"id":"2405.07960","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.07960","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"arxiv_version","alias_value":"2405.07960v5","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.07960","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_12","alias_value":"RRV4GFQO3BSN","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_16","alias_value":"RRV4GFQO3BSNG7R7","created_at":"2026-05-21T13:47:14Z"},{"alias_kind":"pith_short_8","alias_value":"RRV4GFQO","created_at":"2026-05-21T13:47:14Z"}],"graph_snapshots":[{"event_id":"sha256:efede858b0b90b7831d28abaa802b1a3b243960b08e4a228db9f79fdb96d6215","target":"graph","created_at":"2026-05-21T13:47:14Z","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/2405.07960/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating large language models (LLM) in clinical scenarios is crucial to assessing their potential clinical utility. Existing benchmarks rely heavily on static question-answering, which does not accurately depict the complex, sequential nature of clinical decision-making. Here, we introduce AgentClinic, a multimodal agent benchmark for evaluating LLMs in simulated clinical environments that include patient interactions, multimodal data collection under incomplete information, and the usage of various tools, resulting in an in-depth evaluation across nine medical specialties and seven languag","authors_text":"Carl Harris, Eduardo Reis, Jeffrey Jopling, Michael Moor, Rojin Ziaei, Samuel Schmidgall","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-05-13T17:38:53Z","title":"AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environments"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.07960","kind":"arxiv","version":5},"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:e4d12ce6135c05cfdffb1d770b40a88214ec2b860dad4c822a2e673cc70cd468","target":"record","created_at":"2026-05-21T13:47:14Z","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":"6dab0129b3cd64aa7f6b361c04cad632a91ccee45f9bb199d24ecc97c73d5a18","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-05-13T17:38:53Z","title_canon_sha256":"329298b668b7d8094011bd393da2ad3e586fdf5a58433485b44e89dc083faef2"},"schema_version":"1.0","source":{"id":"2405.07960","kind":"arxiv","version":5}},"canonical_sha256":"8c6bc3160ed864d37e3fc8252e892fa5fefc98956feda7b04dfc1e388a6f7a0e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8c6bc3160ed864d37e3fc8252e892fa5fefc98956feda7b04dfc1e388a6f7a0e","first_computed_at":"2026-05-21T13:47:14.138213Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-21T13:47:14.138213Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rQf3bBRBTZsz/4Vr4rxTme1qf+PMR+V9DMPuguzUiFFhoYMqzvhhK5bbEGxVpNBjlpsLFCJG2z1GO9FDLmHSBQ==","signature_status":"signed_v1","signed_at":"2026-05-21T13:47:14.140102Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.07960","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e4d12ce6135c05cfdffb1d770b40a88214ec2b860dad4c822a2e673cc70cd468","sha256:efede858b0b90b7831d28abaa802b1a3b243960b08e4a228db9f79fdb96d6215"],"state_sha256":"813f5a05b72a3f83dcaf36aed8544c1711b246b980f47ae714c6f358cf4b2875"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CTftiNcrZILRwhM9N+O2UKOvirfKYDf8KW0v3UJ6WuFORGcb4adg8UtwflUQ1bdBCEiig1z59jl4zS8Ey6TVCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T03:43:22.334835Z","bundle_sha256":"ad74a92daaf312d915e68f2a0abb3d5a0b61f4c19b742185a3de10bd23d976ae"}}