{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:T5OKJ53A3AZZFEAL2MHOWDLH6N","short_pith_number":"pith:T5OKJ53A","schema_version":"1.0","canonical_sha256":"9f5ca4f760d83392900bd30eeb0d67f354f8d15387e3dff106c7df1af9f8a3d5","source":{"kind":"arxiv","id":"2209.15558","version":2},"attestation_state":"computed","paper":{"title":"Out-of-Distribution Detection and Selective Generation for Conditional Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Balaji Lakshminarayanan, Jiaming Luo, Jie Ren, Kundan Krishna, Mohammad Saleh, Peter J. Liu, Yao Zhao","submitted_at":"2022-09-30T16:17:11Z","abstract_excerpt":"Machine learning algorithms typically assume independent and identically distributed samples in training and at test time. Much work has shown that high-performing ML classifiers can degrade significantly and provide overly-confident, wrong classification predictions, particularly for out-of-distribution (OOD) inputs. Conditional language models (CLMs) are predominantly trained to classify the next token in an output sequence, and may suffer even worse degradation on OOD inputs as the prediction is done auto-regressively over many steps. Furthermore, the space of potential low-quality outputs "},"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":"2209.15558","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-09-30T16:17:11Z","cross_cats_sorted":[],"title_canon_sha256":"24d6cd29c3d96b74fd2161661a40be79e3fcd5c3a53cc109310bb0ac4009dc5a","abstract_canon_sha256":"4f433a95e2a3d17af2325cd129f2ca152f91052a5a5cab6a2cdaa29eb423a275"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:49:04.335060Z","signature_b64":"3k8mbMFuIQxy4lZ4fD5QxovN3HIdc9sOfBcCluH9jgu3Mme+psSDewLHXGMsgjhnw1eqDeG3I3WIUuuw5Bu+AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9f5ca4f760d83392900bd30eeb0d67f354f8d15387e3dff106c7df1af9f8a3d5","last_reissued_at":"2026-07-05T05:49:04.334654Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:49:04.334654Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Out-of-Distribution Detection and Selective Generation for Conditional Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Balaji Lakshminarayanan, Jiaming Luo, Jie Ren, Kundan Krishna, Mohammad Saleh, Peter J. Liu, Yao Zhao","submitted_at":"2022-09-30T16:17:11Z","abstract_excerpt":"Machine learning algorithms typically assume independent and identically distributed samples in training and at test time. Much work has shown that high-performing ML classifiers can degrade significantly and provide overly-confident, wrong classification predictions, particularly for out-of-distribution (OOD) inputs. Conditional language models (CLMs) are predominantly trained to classify the next token in an output sequence, and may suffer even worse degradation on OOD inputs as the prediction is done auto-regressively over many steps. Furthermore, the space of potential low-quality outputs "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.15558","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/2209.15558/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":"2209.15558","created_at":"2026-07-05T05:49:04.334712+00:00"},{"alias_kind":"arxiv_version","alias_value":"2209.15558v2","created_at":"2026-07-05T05:49:04.334712+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.15558","created_at":"2026-07-05T05:49:04.334712+00:00"},{"alias_kind":"pith_short_12","alias_value":"T5OKJ53A3AZZ","created_at":"2026-07-05T05:49:04.334712+00:00"},{"alias_kind":"pith_short_16","alias_value":"T5OKJ53A3AZZFEAL","created_at":"2026-07-05T05:49:04.334712+00:00"},{"alias_kind":"pith_short_8","alias_value":"T5OKJ53A","created_at":"2026-07-05T05:49:04.334712+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.06959","citing_title":"OpenHalDet: A Unified Benchmark for Hallucination Detection across Diverse Generation Scenarios","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01033","citing_title":"TriLens: Per-Layer Logit-Lens Entropy for White-Box Hallucination Detection","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28264","citing_title":"Entropy Distribution as a Fingerprint for Hallucinations in Generative Models","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2602.04572","citing_title":"From Competition to Collaboration: Designing Sustainable Mechanisms Between LLMs and Online Forums","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05638","citing_title":"Scaling Pretrained Representations Enables Label-Free Out-of-Distribution Detection Without Fine-Tuning","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15741","citing_title":"Learning Uncertainty from Sequential Internal Dispersion in Large Language Models","ref_index":38,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N","json":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N.json","graph_json":"https://pith.science/api/pith-number/T5OKJ53A3AZZFEAL2MHOWDLH6N/graph.json","events_json":"https://pith.science/api/pith-number/T5OKJ53A3AZZFEAL2MHOWDLH6N/events.json","paper":"https://pith.science/paper/T5OKJ53A"},"agent_actions":{"view_html":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N","download_json":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N.json","view_paper":"https://pith.science/paper/T5OKJ53A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2209.15558&json=true","fetch_graph":"https://pith.science/api/pith-number/T5OKJ53A3AZZFEAL2MHOWDLH6N/graph.json","fetch_events":"https://pith.science/api/pith-number/T5OKJ53A3AZZFEAL2MHOWDLH6N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N/action/storage_attestation","attest_author":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N/action/author_attestation","sign_citation":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N/action/citation_signature","submit_replication":"https://pith.science/pith/T5OKJ53A3AZZFEAL2MHOWDLH6N/action/replication_record"}},"created_at":"2026-07-05T05:49:04.334712+00:00","updated_at":"2026-07-05T05:49:04.334712+00:00"}