{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:I455A524BATPWTHXFIWSTKOIYM","short_pith_number":"pith:I455A524","canonical_record":{"source":{"id":"2309.05619","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-11T17:07:01Z","cross_cats_sorted":[],"title_canon_sha256":"a95d70d5f3bd410fea2b4c95d28e8bd3d2d7e27306fe5ea973ba9ae2ab1e931c","abstract_canon_sha256":"4a90cfc210198436568aa03565fc8717a108dfd2c84551a652adbc09dfb6ffcc"},"schema_version":"1.0"},"canonical_sha256":"473bd0775c0826fb4cf72a2d29a9c8c3276379b297316671ea914352290e648c","source":{"kind":"arxiv","id":"2309.05619","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.05619","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"arxiv_version","alias_value":"2309.05619v2","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.05619","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_12","alias_value":"I455A524BATP","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_16","alias_value":"I455A524BATPWTHX","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_8","alias_value":"I455A524","created_at":"2026-07-05T07:14:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:I455A524BATPWTHXFIWSTKOIYM","target":"record","payload":{"canonical_record":{"source":{"id":"2309.05619","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-11T17:07:01Z","cross_cats_sorted":[],"title_canon_sha256":"a95d70d5f3bd410fea2b4c95d28e8bd3d2d7e27306fe5ea973ba9ae2ab1e931c","abstract_canon_sha256":"4a90cfc210198436568aa03565fc8717a108dfd2c84551a652adbc09dfb6ffcc"},"schema_version":"1.0"},"canonical_sha256":"473bd0775c0826fb4cf72a2d29a9c8c3276379b297316671ea914352290e648c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:14:17.174975Z","signature_b64":"+1lKTwibXLaucTiLIt3hP5DCnlF4mTYocuSm7JEoUly6qcRDuXEbGvxv+8NEn4W8qSsY5AX5DclrlTb/luIHCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"473bd0775c0826fb4cf72a2d29a9c8c3276379b297316671ea914352290e648c","last_reissued_at":"2026-07-05T07:14:17.174533Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:14:17.174533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2309.05619","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":"2026-07-05T07:14:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LtU8W+ypqHVr00zkV8+PTlpNPO+e/KjPB9ZJ48F4DgszpfmMWpanlv6Drsj3ojL2g+CZnqIB/DafmwaS1bjxAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T18:54:27.634880Z"},"content_sha256":"52f08f935bda20c710fd7b9c8fe1608f18b704408002797e7b15df5666d1e9db","schema_version":"1.0","event_id":"sha256:52f08f935bda20c710fd7b9c8fe1608f18b704408002797e7b15df5666d1e9db"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:I455A524BATPWTHXFIWSTKOIYM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Effective Proxy for Human Labeling: Ensemble Disagreement Scores in Large Language Models for Industrial NLP","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aaron Colak, Daniel J Perry, Laksh Advani, Prashant Shiralkar, Wei Du, Yashmeet Gambhir, Zhengzheng Xing","submitted_at":"2023-09-11T17:07:01Z","abstract_excerpt":"Large language models (LLMs) have demonstrated significant capability to generalize across a large number of NLP tasks. For industry applications, it is imperative to assess the performance of the LLM on unlabeled production data from time to time to validate for a real-world setting. Human labeling to assess model error requires considerable expense and time delay. Here we demonstrate that ensemble disagreement scores work well as a proxy for human labeling for language models in zero-shot, few-shot, and fine-tuned settings, per our evaluation on keyphrase extraction (KPE) task. We measure fi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.05619","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/2309.05619/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":"2026-07-05T07:14:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MgSNNFyXCQ+YN0trt3tA/ifY65FGcWnA158zLlf/RQ4d6HlyQxRvGDhUHuT5HlxPnJRX3vqv4jQkfSxtfBxmDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T18:54:27.635397Z"},"content_sha256":"62316ea7da540edf0b600dc50cd1af7680a68ca0ad4620b8734a087c5460c1c5","schema_version":"1.0","event_id":"sha256:62316ea7da540edf0b600dc50cd1af7680a68ca0ad4620b8734a087c5460c1c5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/I455A524BATPWTHXFIWSTKOIYM/bundle.json","state_url":"https://pith.science/pith/I455A524BATPWTHXFIWSTKOIYM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/I455A524BATPWTHXFIWSTKOIYM/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":"2026-08-19T18:54:27Z","links":{"resolver":"https://pith.science/pith/I455A524BATPWTHXFIWSTKOIYM","bundle":"https://pith.science/pith/I455A524BATPWTHXFIWSTKOIYM/bundle.json","state":"https://pith.science/pith/I455A524BATPWTHXFIWSTKOIYM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/I455A524BATPWTHXFIWSTKOIYM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:I455A524BATPWTHXFIWSTKOIYM","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":"4a90cfc210198436568aa03565fc8717a108dfd2c84551a652adbc09dfb6ffcc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-11T17:07:01Z","title_canon_sha256":"a95d70d5f3bd410fea2b4c95d28e8bd3d2d7e27306fe5ea973ba9ae2ab1e931c"},"schema_version":"1.0","source":{"id":"2309.05619","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.05619","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"arxiv_version","alias_value":"2309.05619v2","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.05619","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_12","alias_value":"I455A524BATP","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_16","alias_value":"I455A524BATPWTHX","created_at":"2026-07-05T07:14:17Z"},{"alias_kind":"pith_short_8","alias_value":"I455A524","created_at":"2026-07-05T07:14:17Z"}],"graph_snapshots":[{"event_id":"sha256:62316ea7da540edf0b600dc50cd1af7680a68ca0ad4620b8734a087c5460c1c5","target":"graph","created_at":"2026-07-05T07:14:17Z","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/2309.05619/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have demonstrated significant capability to generalize across a large number of NLP tasks. For industry applications, it is imperative to assess the performance of the LLM on unlabeled production data from time to time to validate for a real-world setting. Human labeling to assess model error requires considerable expense and time delay. Here we demonstrate that ensemble disagreement scores work well as a proxy for human labeling for language models in zero-shot, few-shot, and fine-tuned settings, per our evaluation on keyphrase extraction (KPE) task. We measure fi","authors_text":"Aaron Colak, Daniel J Perry, Laksh Advani, Prashant Shiralkar, Wei Du, Yashmeet Gambhir, Zhengzheng Xing","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-11T17:07:01Z","title":"Effective Proxy for Human Labeling: Ensemble Disagreement Scores in Large Language Models for Industrial NLP"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.05619","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":"sha256:52f08f935bda20c710fd7b9c8fe1608f18b704408002797e7b15df5666d1e9db","target":"record","created_at":"2026-07-05T07:14:17Z","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":"4a90cfc210198436568aa03565fc8717a108dfd2c84551a652adbc09dfb6ffcc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-11T17:07:01Z","title_canon_sha256":"a95d70d5f3bd410fea2b4c95d28e8bd3d2d7e27306fe5ea973ba9ae2ab1e931c"},"schema_version":"1.0","source":{"id":"2309.05619","kind":"arxiv","version":2}},"canonical_sha256":"473bd0775c0826fb4cf72a2d29a9c8c3276379b297316671ea914352290e648c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"473bd0775c0826fb4cf72a2d29a9c8c3276379b297316671ea914352290e648c","first_computed_at":"2026-07-05T07:14:17.174533Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:14:17.174533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"+1lKTwibXLaucTiLIt3hP5DCnlF4mTYocuSm7JEoUly6qcRDuXEbGvxv+8NEn4W8qSsY5AX5DclrlTb/luIHCg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:14:17.174975Z","signed_message":"canonical_sha256_bytes"},"source_id":"2309.05619","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:52f08f935bda20c710fd7b9c8fe1608f18b704408002797e7b15df5666d1e9db","sha256:62316ea7da540edf0b600dc50cd1af7680a68ca0ad4620b8734a087c5460c1c5"],"state_sha256":"a488bb23c530c384c193e8760066c35644c0b704aa1b289ddb78b5a66afae1f6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RUk7StanMnQwCE8hHq3jqbTEf0yv5ZG8o7XYGLMUWhP+zgoWGiwfp9JVDiUVSqzWuNtJgIhW+/1YD00CIbHOBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T18:54:27.639880Z","bundle_sha256":"fabd27a00b00c6c26b4816c4ba84527d664eef2060e23d1cedb7b0a9f47a5ab8"}}