{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:TONP7GQOK2OJWN7NEK36XNOFOZ","short_pith_number":"pith:TONP7GQO","canonical_record":{"source":{"id":"2311.13735","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-22T23:35:13Z","cross_cats_sorted":[],"title_canon_sha256":"13527e06d7df7f7c6c4a843b4d9a0fbf25c2013a6a605a59c07878a6080bd3c8","abstract_canon_sha256":"65a8b6e1dee7deeea234131669b5d118f35fc4217e78349621f354bf4e9ce11d"},"schema_version":"1.0"},"canonical_sha256":"9b9aff9a0e569c9b37ed22b7ebb5c576648582a5af5471d52df7230e68583b80","source":{"kind":"arxiv","id":"2311.13735","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.13735","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"arxiv_version","alias_value":"2311.13735v1","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.13735","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_12","alias_value":"TONP7GQOK2OJ","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_16","alias_value":"TONP7GQOK2OJWN7N","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_8","alias_value":"TONP7GQO","created_at":"2026-07-05T07:16:07Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:TONP7GQOK2OJWN7NEK36XNOFOZ","target":"record","payload":{"canonical_record":{"source":{"id":"2311.13735","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-22T23:35:13Z","cross_cats_sorted":[],"title_canon_sha256":"13527e06d7df7f7c6c4a843b4d9a0fbf25c2013a6a605a59c07878a6080bd3c8","abstract_canon_sha256":"65a8b6e1dee7deeea234131669b5d118f35fc4217e78349621f354bf4e9ce11d"},"schema_version":"1.0"},"canonical_sha256":"9b9aff9a0e569c9b37ed22b7ebb5c576648582a5af5471d52df7230e68583b80","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:16:07.086928Z","signature_b64":"Mk4LKvjZWpfkyeCymLA69EYroux9zTRLqR9H0ST+hV22/vdo9ISiQorqerRGH3UE6of961mS119ku7FI5IhHCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9b9aff9a0e569c9b37ed22b7ebb5c576648582a5af5471d52df7230e68583b80","last_reissued_at":"2026-07-05T07:16:07.086430Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:16:07.086430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.13735","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":"2026-07-05T07:16:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WflPuiFTprr/SC9AX4DXLUtRS6GW53ndIS03KGKq2ubK6+P/+zNO6Tq7WFtRW/xBCrHL3fxEzu0hImt6kTQTBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T19:38:10.420962Z"},"content_sha256":"1ae80cab738f96a066b49a615417b64173d383bf1a1f1606a5f0c3be92d51671","schema_version":"1.0","event_id":"sha256:1ae80cab738f96a066b49a615417b64173d383bf1a1f1606a5f0c3be92d51671"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:TONP7GQOK2OJWN7NEK36XNOFOZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Surpassing GPT-4 Medical Coding with a Two-Stage Approach","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Eran Halperin, Joel Stremmel, Sanjit Singh Batra, Zhichao Yang","submitted_at":"2023-11-22T23:35:13Z","abstract_excerpt":"Recent advances in large language models (LLMs) show potential for clinical applications, such as clinical decision support and trial recommendations. However, the GPT-4 LLM predicts an excessive number of ICD codes for medical coding tasks, leading to high recall but low precision. To tackle this challenge, we introduce LLM-codex, a two-stage approach to predict ICD codes that first generates evidence proposals using an LLM and then employs an LSTM-based verification stage. The LSTM learns from both the LLM's high recall and human expert's high precision, using a custom loss function. Our mod"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.13735","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/2311.13735/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":"2026-07-05T07:16:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SLlnNpuXBt8Z6xrk46gbYEQ+dfuQJLLU5hZLOAxCOuY/nV+oBHzwHdmbYqMRrJI5IDgcqyCzv2LTOS4HFJZaBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T19:38:10.421473Z"},"content_sha256":"a41dcc096358896b6bc18a667356932734a15c3a890360fc955185c9983784df","schema_version":"1.0","event_id":"sha256:a41dcc096358896b6bc18a667356932734a15c3a890360fc955185c9983784df"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/bundle.json","state_url":"https://pith.science/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/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":"2026-08-07T19:38:10Z","links":{"resolver":"https://pith.science/pith/TONP7GQOK2OJWN7NEK36XNOFOZ","bundle":"https://pith.science/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/bundle.json","state":"https://pith.science/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TONP7GQOK2OJWN7NEK36XNOFOZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:TONP7GQOK2OJWN7NEK36XNOFOZ","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":"65a8b6e1dee7deeea234131669b5d118f35fc4217e78349621f354bf4e9ce11d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-22T23:35:13Z","title_canon_sha256":"13527e06d7df7f7c6c4a843b4d9a0fbf25c2013a6a605a59c07878a6080bd3c8"},"schema_version":"1.0","source":{"id":"2311.13735","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.13735","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"arxiv_version","alias_value":"2311.13735v1","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.13735","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_12","alias_value":"TONP7GQOK2OJ","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_16","alias_value":"TONP7GQOK2OJWN7N","created_at":"2026-07-05T07:16:07Z"},{"alias_kind":"pith_short_8","alias_value":"TONP7GQO","created_at":"2026-07-05T07:16:07Z"}],"graph_snapshots":[{"event_id":"sha256:a41dcc096358896b6bc18a667356932734a15c3a890360fc955185c9983784df","target":"graph","created_at":"2026-07-05T07:16:07Z","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/2311.13735/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in large language models (LLMs) show potential for clinical applications, such as clinical decision support and trial recommendations. However, the GPT-4 LLM predicts an excessive number of ICD codes for medical coding tasks, leading to high recall but low precision. To tackle this challenge, we introduce LLM-codex, a two-stage approach to predict ICD codes that first generates evidence proposals using an LLM and then employs an LSTM-based verification stage. The LSTM learns from both the LLM's high recall and human expert's high precision, using a custom loss function. Our mod","authors_text":"Eran Halperin, Joel Stremmel, Sanjit Singh Batra, Zhichao Yang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-22T23:35:13Z","title":"Surpassing GPT-4 Medical Coding with a Two-Stage Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.13735","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":"sha256:1ae80cab738f96a066b49a615417b64173d383bf1a1f1606a5f0c3be92d51671","target":"record","created_at":"2026-07-05T07:16:07Z","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":"65a8b6e1dee7deeea234131669b5d118f35fc4217e78349621f354bf4e9ce11d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-22T23:35:13Z","title_canon_sha256":"13527e06d7df7f7c6c4a843b4d9a0fbf25c2013a6a605a59c07878a6080bd3c8"},"schema_version":"1.0","source":{"id":"2311.13735","kind":"arxiv","version":1}},"canonical_sha256":"9b9aff9a0e569c9b37ed22b7ebb5c576648582a5af5471d52df7230e68583b80","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9b9aff9a0e569c9b37ed22b7ebb5c576648582a5af5471d52df7230e68583b80","first_computed_at":"2026-07-05T07:16:07.086430Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:16:07.086430Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Mk4LKvjZWpfkyeCymLA69EYroux9zTRLqR9H0ST+hV22/vdo9ISiQorqerRGH3UE6of961mS119ku7FI5IhHCA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:16:07.086928Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.13735","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1ae80cab738f96a066b49a615417b64173d383bf1a1f1606a5f0c3be92d51671","sha256:a41dcc096358896b6bc18a667356932734a15c3a890360fc955185c9983784df"],"state_sha256":"5887ffb88bad14d07104721dc778e18775074dac39448ac32e3e5b352a02f086"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FK3b/IpWT4kDHfz5qijcm9OLmHCy0zLhoOKrjR1LKTgALI8YYmBE/pHqEJ3qml9UhtVMsgWFjPbLeg3MjcdwBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T19:38:10.426004Z","bundle_sha256":"523d84e14c8c607b93cb09cdff65547ca269624e18e24d2de4e39d70e145e83a"}}