{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:O2S4JVHLZ4UV24CCGGGLYHYO62","short_pith_number":"pith:O2S4JVHL","schema_version":"1.0","canonical_sha256":"76a5c4d4ebcf295d7042318cbc1f0ef6a9d6da943518118b18eb4e96ced4522a","source":{"kind":"arxiv","id":"1906.02994","version":2},"attestation_state":"computed","paper":{"title":"Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Akihiro Matsukawa, Balaji Lakshminarayanan, Eric Nalisnick, Yee Whye Teh","submitted_at":"2019-06-07T10:03:16Z","abstract_excerpt":"Recent work has shown that deep generative models can assign higher likelihood to out-of-distribution data sets than to their training data (Nalisnick et al., 2019; Choi et al., 2019). We posit that this phenomenon is caused by a mismatch between the model's typical set and its areas of high probability density. In-distribution inputs should reside in the former but not necessarily in the latter, as previous work has presumed. To determine whether or not inputs reside in the typical set, we propose a statistically principled, easy-to-implement test using the empirical distribution of model lik"},"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":"1906.02994","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2019-06-07T10:03:16Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2e6825e8754bf13043392423188a3485e3d423d91d423deb3c59c3d02056f668","abstract_canon_sha256":"2b624f9ef6766f3238467360f5e767d4891dfa76ee83e75135f71073a9a5c730"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:12:26.617122Z","signature_b64":"P8fZg2UqlRQQ8ayspO938oDBc11p8WmxDvFZNtGgibr56BrLdx/hZ6gZ5oT8jaJ+jOkZu/SzxL+PzgOBTiEFDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76a5c4d4ebcf295d7042318cbc1f0ef6a9d6da943518118b18eb4e96ced4522a","last_reissued_at":"2026-07-05T00:12:26.616705Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:12:26.616705Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Akihiro Matsukawa, Balaji Lakshminarayanan, Eric Nalisnick, Yee Whye Teh","submitted_at":"2019-06-07T10:03:16Z","abstract_excerpt":"Recent work has shown that deep generative models can assign higher likelihood to out-of-distribution data sets than to their training data (Nalisnick et al., 2019; Choi et al., 2019). We posit that this phenomenon is caused by a mismatch between the model's typical set and its areas of high probability density. In-distribution inputs should reside in the former but not necessarily in the latter, as previous work has presumed. To determine whether or not inputs reside in the typical set, we propose a statistically principled, easy-to-implement test using the empirical distribution of model lik"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1906.02994","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/1906.02994/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":"1906.02994","created_at":"2026-07-05T00:12:26.616757+00:00"},{"alias_kind":"arxiv_version","alias_value":"1906.02994v2","created_at":"2026-07-05T00:12:26.616757+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1906.02994","created_at":"2026-07-05T00:12:26.616757+00:00"},{"alias_kind":"pith_short_12","alias_value":"O2S4JVHLZ4UV","created_at":"2026-07-05T00:12:26.616757+00:00"},{"alias_kind":"pith_short_16","alias_value":"O2S4JVHLZ4UV24CC","created_at":"2026-07-05T00:12:26.616757+00:00"},{"alias_kind":"pith_short_8","alias_value":"O2S4JVHL","created_at":"2026-07-05T00:12:26.616757+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08417","citing_title":"Hacking Generative Perplexity: Why Unconditional Text Evaluation Needs Distributional Metrics","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00267","citing_title":"StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31596","citing_title":"KLIP: localized distribution shift detection via KL-divergence with diffusion priors in Inverse Problems","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17034","citing_title":"Privacy Policy Enforcement Guardrails for Data-Sensitive Retrieval-Augmented Generation","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2502.11638","citing_title":"Safeguarding AI in Medical Imaging: Post-Hoc Out-of-Distribution Detection with Normalizing Flows","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22496","citing_title":"The Signal in the Noise: OOD Detection Through Goodness-of-Fit Testing in Factorised Latent Spaces","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17034","citing_title":"Privacy Policy Enforcement Guardrails for Data-Sensitive Retrieval-Augmented Generation","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03802","citing_title":"Towards accurate extreme event likelihoods from diffusion model climate emulators","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08202","citing_title":"Beyond Penalization: Diffusion-based Out-of-Distribution Detection and Selective Regularization in Offline Reinforcement Learning","ref_index":34,"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":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62","json":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62.json","graph_json":"https://pith.science/api/pith-number/O2S4JVHLZ4UV24CCGGGLYHYO62/graph.json","events_json":"https://pith.science/api/pith-number/O2S4JVHLZ4UV24CCGGGLYHYO62/events.json","paper":"https://pith.science/paper/O2S4JVHL"},"agent_actions":{"view_html":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62","download_json":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62.json","view_paper":"https://pith.science/paper/O2S4JVHL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1906.02994&json=true","fetch_graph":"https://pith.science/api/pith-number/O2S4JVHLZ4UV24CCGGGLYHYO62/graph.json","fetch_events":"https://pith.science/api/pith-number/O2S4JVHLZ4UV24CCGGGLYHYO62/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62/action/storage_attestation","attest_author":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62/action/author_attestation","sign_citation":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62/action/citation_signature","submit_replication":"https://pith.science/pith/O2S4JVHLZ4UV24CCGGGLYHYO62/action/replication_record"}},"created_at":"2026-07-05T00:12:26.616757+00:00","updated_at":"2026-07-05T00:12:26.616757+00:00"}