{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:WS5ZVAQYER47LYIIATKZTFIHEI","short_pith_number":"pith:WS5ZVAQY","canonical_record":{"source":{"id":"2406.00922","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T01:32:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"79503d5940377b18bad3bde2b6d42c2b27059502e34b3838bc49daac72caa59d","abstract_canon_sha256":"cb03f5bfc21e40f975a27ae61d5b5c60fce9ed1207e0ba09bf755906629d48d5"},"schema_version":"1.0"},"canonical_sha256":"b4bb9a82182479f5e10804d5999507222e843971e46c433767339734faf8aa87","source":{"kind":"arxiv","id":"2406.00922","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.00922","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"arxiv_version","alias_value":"2406.00922v3","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.00922","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_12","alias_value":"WS5ZVAQYER47","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_16","alias_value":"WS5ZVAQYER47LYII","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_8","alias_value":"WS5ZVAQY","created_at":"2026-07-05T09:32:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:WS5ZVAQYER47LYIIATKZTFIHEI","target":"record","payload":{"canonical_record":{"source":{"id":"2406.00922","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T01:32:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"79503d5940377b18bad3bde2b6d42c2b27059502e34b3838bc49daac72caa59d","abstract_canon_sha256":"cb03f5bfc21e40f975a27ae61d5b5c60fce9ed1207e0ba09bf755906629d48d5"},"schema_version":"1.0"},"canonical_sha256":"b4bb9a82182479f5e10804d5999507222e843971e46c433767339734faf8aa87","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:32:31.964258Z","signature_b64":"XUmAVaiKiPgf9riRbWycxeuVYMsGGO0m92axlOCYiw+FxKo+metQIKcdbVvHKGjr4bf5W3iH3p5DKy1j1zLxAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b4bb9a82182479f5e10804d5999507222e843971e46c433767339734faf8aa87","last_reissued_at":"2026-07-05T09:32:31.963748Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:32:31.963748Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.00922","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:32:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LfvYViZYKCeHKy+TwFBM8Q3r7ppf5fGbaX77BCItcT7V/eNFqjfeJ67vOyQI++mZFNyAXFkqokW7reKUnut6Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T20:56:50.007863Z"},"content_sha256":"d1b9102c5942b0d3e1d4a4bdd08c98eab5c22605931a43d29979f49ed45088b6","schema_version":"1.0","event_id":"sha256:d1b9102c5942b0d3e1d4a4bdd08c98eab5c22605931a43d29979f49ed45088b6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:WS5ZVAQYER47LYIIATKZTFIHEI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MediQ: Question-Asking LLMs and a Benchmark for Reliable Interactive Clinical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Emma Pierson, Jonathan S. Ilgen, Pang Wei Koh, Shangbin Feng, Shuyue Stella Li, Vidhisha Balachandran, Yulia Tsvetkov","submitted_at":"2024-06-03T01:32:52Z","abstract_excerpt":"Users typically engage with LLMs interactively, yet most existing benchmarks evaluate them in a static, single-turn format, posing reliability concerns in interactive scenarios. We identify a key obstacle towards reliability: LLMs are trained to answer any question, even with incomplete context or insufficient knowledge. In this paper, we propose to change the static paradigm to an interactive one, develop systems that proactively ask questions to gather more information and respond reliably, and introduce an benchmark - MediQ - to evaluate question-asking ability in LLMs. MediQ simulates clin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.00922","kind":"arxiv","version":3},"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/2406.00922/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":"2026-07-05T09:32:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4UaixRUt9ojay/8dTLe/9nGGe0ol8WYJ4ASvXv9J2iVbI4/nyRHZJTS5MhBejeYCSmbTFzxZ+oCk6VQOpqxfCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T20:56:50.008261Z"},"content_sha256":"1e2183c9c55027adddaaa70c2cace87cd5f6aaf4576e034ff50e5c30ce9134fc","schema_version":"1.0","event_id":"sha256:1e2183c9c55027adddaaa70c2cace87cd5f6aaf4576e034ff50e5c30ce9134fc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WS5ZVAQYER47LYIIATKZTFIHEI/bundle.json","state_url":"https://pith.science/pith/WS5ZVAQYER47LYIIATKZTFIHEI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WS5ZVAQYER47LYIIATKZTFIHEI/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":"2026-07-06T20:56:50Z","links":{"resolver":"https://pith.science/pith/WS5ZVAQYER47LYIIATKZTFIHEI","bundle":"https://pith.science/pith/WS5ZVAQYER47LYIIATKZTFIHEI/bundle.json","state":"https://pith.science/pith/WS5ZVAQYER47LYIIATKZTFIHEI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WS5ZVAQYER47LYIIATKZTFIHEI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:WS5ZVAQYER47LYIIATKZTFIHEI","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":"cb03f5bfc21e40f975a27ae61d5b5c60fce9ed1207e0ba09bf755906629d48d5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T01:32:52Z","title_canon_sha256":"79503d5940377b18bad3bde2b6d42c2b27059502e34b3838bc49daac72caa59d"},"schema_version":"1.0","source":{"id":"2406.00922","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.00922","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"arxiv_version","alias_value":"2406.00922v3","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.00922","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_12","alias_value":"WS5ZVAQYER47","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_16","alias_value":"WS5ZVAQYER47LYII","created_at":"2026-07-05T09:32:31Z"},{"alias_kind":"pith_short_8","alias_value":"WS5ZVAQY","created_at":"2026-07-05T09:32:31Z"}],"graph_snapshots":[{"event_id":"sha256:1e2183c9c55027adddaaa70c2cace87cd5f6aaf4576e034ff50e5c30ce9134fc","target":"graph","created_at":"2026-07-05T09:32:31Z","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/2406.00922/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Users typically engage with LLMs interactively, yet most existing benchmarks evaluate them in a static, single-turn format, posing reliability concerns in interactive scenarios. We identify a key obstacle towards reliability: LLMs are trained to answer any question, even with incomplete context or insufficient knowledge. In this paper, we propose to change the static paradigm to an interactive one, develop systems that proactively ask questions to gather more information and respond reliably, and introduce an benchmark - MediQ - to evaluate question-asking ability in LLMs. MediQ simulates clin","authors_text":"Emma Pierson, Jonathan S. Ilgen, Pang Wei Koh, Shangbin Feng, Shuyue Stella Li, Vidhisha Balachandran, Yulia Tsvetkov","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T01:32:52Z","title":"MediQ: Question-Asking LLMs and a Benchmark for Reliable Interactive Clinical Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.00922","kind":"arxiv","version":3},"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:d1b9102c5942b0d3e1d4a4bdd08c98eab5c22605931a43d29979f49ed45088b6","target":"record","created_at":"2026-07-05T09:32:31Z","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":"cb03f5bfc21e40f975a27ae61d5b5c60fce9ed1207e0ba09bf755906629d48d5","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T01:32:52Z","title_canon_sha256":"79503d5940377b18bad3bde2b6d42c2b27059502e34b3838bc49daac72caa59d"},"schema_version":"1.0","source":{"id":"2406.00922","kind":"arxiv","version":3}},"canonical_sha256":"b4bb9a82182479f5e10804d5999507222e843971e46c433767339734faf8aa87","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b4bb9a82182479f5e10804d5999507222e843971e46c433767339734faf8aa87","first_computed_at":"2026-07-05T09:32:31.963748Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:32:31.963748Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"XUmAVaiKiPgf9riRbWycxeuVYMsGGO0m92axlOCYiw+FxKo+metQIKcdbVvHKGjr4bf5W3iH3p5DKy1j1zLxAw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:32:31.964258Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.00922","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d1b9102c5942b0d3e1d4a4bdd08c98eab5c22605931a43d29979f49ed45088b6","sha256:1e2183c9c55027adddaaa70c2cace87cd5f6aaf4576e034ff50e5c30ce9134fc"],"state_sha256":"ebf69d155478dd615467d7cf8ef78392a92604aa1e19e9f689836a2a4ea44ea6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QAd00D4qoRQWnhS9EMCxnK+E3sZMDzfHzy2TGodU3W1Ibv91FcmgheISbYHGPYaw5aF+wKyqZ5niyHWEDeoTCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T20:56:50.010292Z","bundle_sha256":"2cec41a504591aa03500586216b7bab33cdc66bb491be5e91fa4cfcb04374dc3"}}