{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:R2MRYGUI647XIER6PIVINYT4C3","short_pith_number":"pith:R2MRYGUI","schema_version":"1.0","canonical_sha256":"8e991c1a88f73f74123e7a2a86e27c16e5eeb4468b8054bb8c4b54e1d0c92f14","source":{"kind":"arxiv","id":"2401.08703","version":2},"attestation_state":"computed","paper":{"title":"Decoupled Prototype Learning for Reliable Test-Time Adaptation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Changxing Ding, Guowei Wang, Mingkui Tan, Wentao Tan","submitted_at":"2024-01-15T03:33:39Z","abstract_excerpt":"Test-time adaptation (TTA) is a task that continually adapts a pre-trained source model to the target domain during inference. One popular approach involves fine-tuning model with cross-entropy loss according to estimated pseudo-labels. However, its performance is significantly affected by noisy pseudo-labels. This study reveals that minimizing the classification error of each sample causes the cross-entropy loss's vulnerability to label noise. To address this issue, we propose a novel Decoupled Prototype Learning (DPL) method that features prototype-centric loss computation. First, we decoupl"},"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":"2401.08703","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-15T03:33:39Z","cross_cats_sorted":[],"title_canon_sha256":"aa796cbe647e3bb8320e175132f9a6d6cba5210093d2575d0d11febc5e446c3d","abstract_canon_sha256":"4ac87a35d9a3bfc46cdb00b8b19455fd016b210d8c15ebf5ab82ca755a823f9a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:37:15.170698Z","signature_b64":"OczbVkSM9UxGldU5SYz+svNGeeW2ZgwvKOfpklR538bhPRNc52d4ViM0R8+DhZeMzDMI6R1251+R/Vt6YlvPBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8e991c1a88f73f74123e7a2a86e27c16e5eeb4468b8054bb8c4b54e1d0c92f14","last_reissued_at":"2026-07-05T07:37:15.170231Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:37:15.170231Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Decoupled Prototype Learning for Reliable Test-Time Adaptation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Changxing Ding, Guowei Wang, Mingkui Tan, Wentao Tan","submitted_at":"2024-01-15T03:33:39Z","abstract_excerpt":"Test-time adaptation (TTA) is a task that continually adapts a pre-trained source model to the target domain during inference. One popular approach involves fine-tuning model with cross-entropy loss according to estimated pseudo-labels. However, its performance is significantly affected by noisy pseudo-labels. This study reveals that minimizing the classification error of each sample causes the cross-entropy loss's vulnerability to label noise. To address this issue, we propose a novel Decoupled Prototype Learning (DPL) method that features prototype-centric loss computation. First, we decoupl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.08703","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/2401.08703/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":"2401.08703","created_at":"2026-07-05T07:37:15.170288+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.08703v2","created_at":"2026-07-05T07:37:15.170288+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.08703","created_at":"2026-07-05T07:37:15.170288+00:00"},{"alias_kind":"pith_short_12","alias_value":"R2MRYGUI647X","created_at":"2026-07-05T07:37:15.170288+00:00"},{"alias_kind":"pith_short_16","alias_value":"R2MRYGUI647XIER6","created_at":"2026-07-05T07:37:15.170288+00:00"},{"alias_kind":"pith_short_8","alias_value":"R2MRYGUI","created_at":"2026-07-05T07:37:15.170288+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.06782","citing_title":"Feature-Based Instance Neighbor Discovery: Advanced Stable Test-Time Adaptation in Dynamic World","ref_index":27,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3","json":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3.json","graph_json":"https://pith.science/api/pith-number/R2MRYGUI647XIER6PIVINYT4C3/graph.json","events_json":"https://pith.science/api/pith-number/R2MRYGUI647XIER6PIVINYT4C3/events.json","paper":"https://pith.science/paper/R2MRYGUI"},"agent_actions":{"view_html":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3","download_json":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3.json","view_paper":"https://pith.science/paper/R2MRYGUI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.08703&json=true","fetch_graph":"https://pith.science/api/pith-number/R2MRYGUI647XIER6PIVINYT4C3/graph.json","fetch_events":"https://pith.science/api/pith-number/R2MRYGUI647XIER6PIVINYT4C3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3/action/storage_attestation","attest_author":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3/action/author_attestation","sign_citation":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3/action/citation_signature","submit_replication":"https://pith.science/pith/R2MRYGUI647XIER6PIVINYT4C3/action/replication_record"}},"created_at":"2026-07-05T07:37:15.170288+00:00","updated_at":"2026-07-05T07:37:15.170288+00:00"}