{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UXUJNLDHYLUGNQERUF35KWE2LO","short_pith_number":"pith:UXUJNLDH","schema_version":"1.0","canonical_sha256":"a5e896ac67c2e866c091a177d5589a5b88d4ed050cd2811f78c8f8ffb03bdbb4","source":{"kind":"arxiv","id":"2506.02843","version":1},"attestation_state":"computed","paper":{"title":"Random Registers for Cross-Domain Few-Shot Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ruixuan Li, Shuai Yi, Yixiong Zou, Yuhua Li","submitted_at":"2025-06-03T13:13:58Z","abstract_excerpt":"Cross-domain few-shot learning (CDFSL) aims to transfer knowledge from a data-sufficient source domain to data-scarce target domains. Although Vision Transformer (ViT) has shown superior capability in many vision tasks, its transferability against huge domain gaps in CDFSL is still under-explored. In this paper, we find an intriguing phenomenon: during the source-domain training, prompt tuning, as a common way to train ViT, could be harmful for the generalization of ViT in target domains, but setting them to random noises (i.e., random registers) could consistently improve target-domain perfor"},"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":"2506.02843","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-06-03T13:13:58Z","cross_cats_sorted":[],"title_canon_sha256":"d92a846f09968936f3976daea2b29416b3f72887000dbda3f5d12e59fa15eb28","abstract_canon_sha256":"60842e000262455d9e7eb326f4dbe543845db3c7c12fc160a20b741b7f162b65"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:15:10.271908Z","signature_b64":"xKt+uWWuULcYnFhxGaZIKhfIbpguOdkICs7h5xJguVgIpIM2OMjdNJ25y3c1uextUww8vw22oxnqZl5huQ9/BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a5e896ac67c2e866c091a177d5589a5b88d4ed050cd2811f78c8f8ffb03bdbb4","last_reissued_at":"2026-07-05T11:15:10.271372Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:15:10.271372Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Random Registers for Cross-Domain Few-Shot Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ruixuan Li, Shuai Yi, Yixiong Zou, Yuhua Li","submitted_at":"2025-06-03T13:13:58Z","abstract_excerpt":"Cross-domain few-shot learning (CDFSL) aims to transfer knowledge from a data-sufficient source domain to data-scarce target domains. Although Vision Transformer (ViT) has shown superior capability in many vision tasks, its transferability against huge domain gaps in CDFSL is still under-explored. In this paper, we find an intriguing phenomenon: during the source-domain training, prompt tuning, as a common way to train ViT, could be harmful for the generalization of ViT in target domains, but setting them to random noises (i.e., random registers) could consistently improve target-domain perfor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.02843","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/2506.02843/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":"2506.02843","created_at":"2026-07-05T11:15:10.271437+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.02843v1","created_at":"2026-07-05T11:15:10.271437+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.02843","created_at":"2026-07-05T11:15:10.271437+00:00"},{"alias_kind":"pith_short_12","alias_value":"UXUJNLDHYLUG","created_at":"2026-07-05T11:15:10.271437+00:00"},{"alias_kind":"pith_short_16","alias_value":"UXUJNLDHYLUGNQER","created_at":"2026-07-05T11:15:10.271437+00:00"},{"alias_kind":"pith_short_8","alias_value":"UXUJNLDH","created_at":"2026-07-05T11:15:10.271437+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.25799","citing_title":"Addressing Exacerbated Attention Sink for Source-Free Cross-Domain Few-Shot Learning","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29776","citing_title":"Improving CLIP Adaptation by Breaking Tail Alignment for Source-Free Cross-Domain Few-Shot Learning","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO","json":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO.json","graph_json":"https://pith.science/api/pith-number/UXUJNLDHYLUGNQERUF35KWE2LO/graph.json","events_json":"https://pith.science/api/pith-number/UXUJNLDHYLUGNQERUF35KWE2LO/events.json","paper":"https://pith.science/paper/UXUJNLDH"},"agent_actions":{"view_html":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO","download_json":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO.json","view_paper":"https://pith.science/paper/UXUJNLDH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.02843&json=true","fetch_graph":"https://pith.science/api/pith-number/UXUJNLDHYLUGNQERUF35KWE2LO/graph.json","fetch_events":"https://pith.science/api/pith-number/UXUJNLDHYLUGNQERUF35KWE2LO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO/action/storage_attestation","attest_author":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO/action/author_attestation","sign_citation":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO/action/citation_signature","submit_replication":"https://pith.science/pith/UXUJNLDHYLUGNQERUF35KWE2LO/action/replication_record"}},"created_at":"2026-07-05T11:15:10.271437+00:00","updated_at":"2026-07-05T11:15:10.271437+00:00"}