{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ZRLP2JA3DDKQB53KNFKUJBBZH7","short_pith_number":"pith:ZRLP2JA3","schema_version":"1.0","canonical_sha256":"cc56fd241b18d500f76a69554484393fc022285dfed21f6f27098263c417fe6a","source":{"kind":"arxiv","id":"2607.03900","version":1},"attestation_state":"computed","paper":{"title":"USE: A Unified Self-Ensembling Framework for Test-Time Prompt Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jian Liang, Ran He, Siru Jiang, Tieniu Tan","submitted_at":"2026-07-04T14:45:23Z","abstract_excerpt":"Test-time adaptation (TTA) has emerged as a popular paradigm for improving the performance of vision-language models (e.g., CLIP) on downstream tasks. Among existing CLIP-based TTA methods, Test-Time Prompt Tuning (TPT) is a pioneering work that optimizes textual prompts using multiple test-time augmentations and remains a strong baseline to date. In this work, we revisit TPT and reveal that its optimization can be interpreted as implicitly learning from self-generated pseudo labels. Building on this perspective, we propose a unified self-ensembling framework (USE) that ensures consistency bet"},"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":"2607.03900","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-07-04T14:45:23Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"90ad8b4f51bd19b9f8cdfe46d719feb15646789886e6e83e793ef95d0d2043b2","abstract_canon_sha256":"dc0584f81282549de891fdf6f6fb5d3fe717154b418f07d8305649c1e2806866"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:18:13.604285Z","signature_b64":"Z0Ib6ueQUUzNr3x/V2jlYstt9dQzumG+UCBQbNaz4xRkUePPPY1U6T2niBs6Nnhr4UL5pfFgdRtepLc9vzCQBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cc56fd241b18d500f76a69554484393fc022285dfed21f6f27098263c417fe6a","last_reissued_at":"2026-07-07T02:18:13.603148Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:18:13.603148Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"USE: A Unified Self-Ensembling Framework for Test-Time Prompt Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Jian Liang, Ran He, Siru Jiang, Tieniu Tan","submitted_at":"2026-07-04T14:45:23Z","abstract_excerpt":"Test-time adaptation (TTA) has emerged as a popular paradigm for improving the performance of vision-language models (e.g., CLIP) on downstream tasks. Among existing CLIP-based TTA methods, Test-Time Prompt Tuning (TPT) is a pioneering work that optimizes textual prompts using multiple test-time augmentations and remains a strong baseline to date. In this work, we revisit TPT and reveal that its optimization can be interpreted as implicitly learning from self-generated pseudo labels. Building on this perspective, we propose a unified self-ensembling framework (USE) that ensures consistency bet"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.03900","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/2607.03900/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":"2607.03900","created_at":"2026-07-07T02:18:13.603373+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.03900v1","created_at":"2026-07-07T02:18:13.603373+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.03900","created_at":"2026-07-07T02:18:13.603373+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZRLP2JA3DDKQ","created_at":"2026-07-07T02:18:13.603373+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZRLP2JA3DDKQB53K","created_at":"2026-07-07T02:18:13.603373+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZRLP2JA3","created_at":"2026-07-07T02:18:13.603373+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7","json":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7.json","graph_json":"https://pith.science/api/pith-number/ZRLP2JA3DDKQB53KNFKUJBBZH7/graph.json","events_json":"https://pith.science/api/pith-number/ZRLP2JA3DDKQB53KNFKUJBBZH7/events.json","paper":"https://pith.science/paper/ZRLP2JA3"},"agent_actions":{"view_html":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7","download_json":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7.json","view_paper":"https://pith.science/paper/ZRLP2JA3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.03900&json=true","fetch_graph":"https://pith.science/api/pith-number/ZRLP2JA3DDKQB53KNFKUJBBZH7/graph.json","fetch_events":"https://pith.science/api/pith-number/ZRLP2JA3DDKQB53KNFKUJBBZH7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7/action/storage_attestation","attest_author":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7/action/author_attestation","sign_citation":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7/action/citation_signature","submit_replication":"https://pith.science/pith/ZRLP2JA3DDKQB53KNFKUJBBZH7/action/replication_record"}},"created_at":"2026-07-07T02:18:13.603373+00:00","updated_at":"2026-07-07T02:18:13.603373+00:00"}