{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:LLQHMVSOW2ZBQR5LWOVD7IG5JT","short_pith_number":"pith:LLQHMVSO","schema_version":"1.0","canonical_sha256":"5ae076564eb6b21847abb3aa3fa0dd4cec4b284192d8b0b8a2436756cc32efbd","source":{"kind":"arxiv","id":"1912.02781","version":2},"attestation_state":"computed","paper":{"title":"AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Balaji Lakshminarayanan, Barret Zoph, Dan Hendrycks, Ekin D. Cubuk, Justin Gilmer, Norman Mu","submitted_at":"2019-12-05T18:18:10Z","abstract_excerpt":"Modern deep neural networks can achieve high accuracy when the training distribution and test distribution are identically distributed, but this assumption is frequently violated in practice. When the train and test distributions are mismatched, accuracy can plummet. Currently there are few techniques that improve robustness to unforeseen data shifts encountered during deployment. In this work, we propose a technique to improve the robustness and uncertainty estimates of image classifiers. We propose AugMix, a data processing technique that is simple to implement, adds limited computational ov"},"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":"1912.02781","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2019-12-05T18:18:10Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"f2b7f5c65a7deff724640bee5c59950ed9563367c1948fc7d58f6119dacb464f","abstract_canon_sha256":"f047dc5e9c7bf2e30bccfd538fb7106c1c8430d17f57b12a5f48f9bbf51d25ef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:41:01.283329Z","signature_b64":"RkxQAO7ak4SpceHL96TEyArLTwAFk4IHYo7P1UO41zJJgt/MX7ChtAQaJzrbFTmopZoXG+iFIEtBKNFepfk1Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5ae076564eb6b21847abb3aa3fa0dd4cec4b284192d8b0b8a2436756cc32efbd","last_reissued_at":"2026-07-05T00:41:01.282840Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:41:01.282840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"stat.ML","authors_text":"Balaji Lakshminarayanan, Barret Zoph, Dan Hendrycks, Ekin D. Cubuk, Justin Gilmer, Norman Mu","submitted_at":"2019-12-05T18:18:10Z","abstract_excerpt":"Modern deep neural networks can achieve high accuracy when the training distribution and test distribution are identically distributed, but this assumption is frequently violated in practice. When the train and test distributions are mismatched, accuracy can plummet. Currently there are few techniques that improve robustness to unforeseen data shifts encountered during deployment. In this work, we propose a technique to improve the robustness and uncertainty estimates of image classifiers. We propose AugMix, a data processing technique that is simple to implement, adds limited computational ov"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.02781","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/1912.02781/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":"1912.02781","created_at":"2026-07-05T00:41:01.282901+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.02781v2","created_at":"2026-07-05T00:41:01.282901+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.02781","created_at":"2026-07-05T00:41:01.282901+00:00"},{"alias_kind":"pith_short_12","alias_value":"LLQHMVSOW2ZB","created_at":"2026-07-05T00:41:01.282901+00:00"},{"alias_kind":"pith_short_16","alias_value":"LLQHMVSOW2ZBQR5L","created_at":"2026-07-05T00:41:01.282901+00:00"},{"alias_kind":"pith_short_8","alias_value":"LLQHMVSO","created_at":"2026-07-05T00:41:01.282901+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21736","citing_title":"Adversarial Domain Prompt Tuning and Generation for Single Domain Generalization","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11319","citing_title":"Learning from almost nothing: How neural networks survive heavy input corruption","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06020","citing_title":"ReSAGE-PAR: Representational Similarity Assessment for Generative Expansion in Pedestrian Attribute Recognition","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05536","citing_title":"Dual Feature Decoupling for Fine-Grained OOD Detection","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24508","citing_title":"FDDet: Achieving Data-Efficient Food Defect Detection Under Real-World Scenarios","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21590","citing_title":"Radial Basis Function Networks as Projection Heads in Self-Supervised Learning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2503.02170","citing_title":"Adaptive Camera Sensor for Vision Models","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11558","citing_title":"A Composite Activation Function for Learning Stable Binary Representations","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10756","citing_title":"TINS: Test-time ID-prototype-separated Negative Semantics Learning for OOD Detection","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09716","citing_title":"Medical Model Synthesis Architectures: A Case Study","ref_index":215,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06522","citing_title":"Agentic AIs Are the Missing Paradigm for Out-of-Distribution Generalization in Foundation Models","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20307","citing_title":"Improving Facial Emotion Recognition through Dataset Merging and Balanced Training Strategies","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05583","citing_title":"WRF4CIR: Weight-Regularized Fine-Tuning Network for Composed Image Retrieval","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04488","citing_title":"A Patch-based Cross-view Regularized Framework for Backdoor Defense in Multimodal Large Language Models","ref_index":70,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT","json":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT.json","graph_json":"https://pith.science/api/pith-number/LLQHMVSOW2ZBQR5LWOVD7IG5JT/graph.json","events_json":"https://pith.science/api/pith-number/LLQHMVSOW2ZBQR5LWOVD7IG5JT/events.json","paper":"https://pith.science/paper/LLQHMVSO"},"agent_actions":{"view_html":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT","download_json":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT.json","view_paper":"https://pith.science/paper/LLQHMVSO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.02781&json=true","fetch_graph":"https://pith.science/api/pith-number/LLQHMVSOW2ZBQR5LWOVD7IG5JT/graph.json","fetch_events":"https://pith.science/api/pith-number/LLQHMVSOW2ZBQR5LWOVD7IG5JT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT/action/storage_attestation","attest_author":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT/action/author_attestation","sign_citation":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT/action/citation_signature","submit_replication":"https://pith.science/pith/LLQHMVSOW2ZBQR5LWOVD7IG5JT/action/replication_record"}},"created_at":"2026-07-05T00:41:01.282901+00:00","updated_at":"2026-07-05T00:41:01.282901+00:00"}