{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:UZM6WFZRU3Q7LAOSVCEJI2IR57","short_pith_number":"pith:UZM6WFZR","schema_version":"1.0","canonical_sha256":"a659eb1731a6e1f581d2a888946911efe8864dad230072a99a39119ab13fbefe","source":{"kind":"arxiv","id":"2211.06519","version":1},"attestation_state":"computed","paper":{"title":"The Expertise Problem: Learning from Specialized Feedback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Oliver Daniels-Koch, Rachel Freedman","submitted_at":"2022-11-12T00:07:35Z","abstract_excerpt":"Reinforcement learning from human feedback (RLHF) is a powerful technique for training agents to perform difficult-to-specify tasks. However, human feedback can be noisy, particularly when human teachers lack relevant knowledge or experience. Levels of expertise vary across teachers, and a given teacher may have differing levels of expertise for different components of a task. RLHF algorithms that learn from multiple teachers therefore face an expertise problem: the reliability of a given piece of feedback depends both on the teacher that it comes from and how specialized that teacher is on re"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2211.06519","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-11-12T00:07:35Z","cross_cats_sorted":[],"title_canon_sha256":"64d7f36205b17c8f04bb2e7f4b7b25320c2bfc64ca9fdb63611ca11f9fe11146","abstract_canon_sha256":"fffb86915177130fbfa18e415c33ba3b383cb796c15717d5ce3864e76d3770d3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:15:25.948861Z","signature_b64":"kyWSqxFsJivXDUqoq9oyghVU5si10p2yuqe1f3J8+JkB5y59B4xPxMsK1A+5C0m+12/Frnc8fP1G+IBNT7gVCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a659eb1731a6e1f581d2a888946911efe8864dad230072a99a39119ab13fbefe","last_reissued_at":"2026-07-05T05:15:25.948396Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:15:25.948396Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Expertise Problem: Learning from Specialized Feedback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Oliver Daniels-Koch, Rachel Freedman","submitted_at":"2022-11-12T00:07:35Z","abstract_excerpt":"Reinforcement learning from human feedback (RLHF) is a powerful technique for training agents to perform difficult-to-specify tasks. However, human feedback can be noisy, particularly when human teachers lack relevant knowledge or experience. Levels of expertise vary across teachers, and a given teacher may have differing levels of expertise for different components of a task. RLHF algorithms that learn from multiple teachers therefore face an expertise problem: the reliability of a given piece of feedback depends both on the teacher that it comes from and how specialized that teacher is on re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2211.06519","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/2211.06519/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2211.06519","created_at":"2026-07-05T05:15:25.948455+00:00"},{"alias_kind":"arxiv_version","alias_value":"2211.06519v1","created_at":"2026-07-05T05:15:25.948455+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2211.06519","created_at":"2026-07-05T05:15:25.948455+00:00"},{"alias_kind":"pith_short_12","alias_value":"UZM6WFZRU3Q7","created_at":"2026-07-05T05:15:25.948455+00:00"},{"alias_kind":"pith_short_16","alias_value":"UZM6WFZRU3Q7LAOS","created_at":"2026-07-05T05:15:25.948455+00:00"},{"alias_kind":"pith_short_8","alias_value":"UZM6WFZR","created_at":"2026-07-05T05:15:25.948455+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.16475","citing_title":"When Can Proxies Improve the Sample Complexity of Preference Learning?","ref_index":12,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57","json":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57.json","graph_json":"https://pith.science/api/pith-number/UZM6WFZRU3Q7LAOSVCEJI2IR57/graph.json","events_json":"https://pith.science/api/pith-number/UZM6WFZRU3Q7LAOSVCEJI2IR57/events.json","paper":"https://pith.science/paper/UZM6WFZR"},"agent_actions":{"view_html":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57","download_json":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57.json","view_paper":"https://pith.science/paper/UZM6WFZR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2211.06519&json=true","fetch_graph":"https://pith.science/api/pith-number/UZM6WFZRU3Q7LAOSVCEJI2IR57/graph.json","fetch_events":"https://pith.science/api/pith-number/UZM6WFZRU3Q7LAOSVCEJI2IR57/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57/action/storage_attestation","attest_author":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57/action/author_attestation","sign_citation":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57/action/citation_signature","submit_replication":"https://pith.science/pith/UZM6WFZRU3Q7LAOSVCEJI2IR57/action/replication_record"}},"created_at":"2026-07-05T05:15:25.948455+00:00","updated_at":"2026-07-05T05:15:25.948455+00:00"}