{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7CKSU64KJW7ULWT644BE7447BD","short_pith_number":"pith:7CKSU64K","schema_version":"1.0","canonical_sha256":"f8952a7b8a4dbf45da7ee7024ff39f08d2bb9badb2a808af4629886b18748766","source":{"kind":"arxiv","id":"2608.04935","version":1},"attestation_state":"computed","paper":{"title":"Unleashing the Potential of Vision-Language Models for Generalizable AI-Generated Image Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hao Tan, Jun Wan, Weihan Cai, Xinping Gao, Zichang Tan","submitted_at":"2026-08-05T15:03:44Z","abstract_excerpt":"Recent work has shown that a simple linear probe on frozen representations from modern vision foundation models (VFMs) can achieve state-of-the-art AIGI detection performance, substantially outperforming specialized detectors in challenging in-the-wild scenarios. This finding has established DINOv3 as the dominant foundation-model baseline for subsequent improvements. However, we find that the vision-language model Perception Encoder (PE) holds greater potential for AIGI detection, because its language-aligned representation preserves high-level provenance semantics. Specifically, PE exhibits "},"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":"2608.04935","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-08-05T15:03:44Z","cross_cats_sorted":[],"title_canon_sha256":"e86bd26c8ccd0f4ab0cc27d55fc63411de65a1031cdb15370e2f1954cbe1cb99","abstract_canon_sha256":"c5b2d12063d62c91c7c7157bd4ea088bdf461a7c8f5573195bddfa3f45082894"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:47:43.964660Z","signature_b64":"qjVmQ5qBVW2nOliVWkSmvew1VPEOxh9sraQKABxDNX95L01REmM1MYqKxpgoHUFDIf5DyN3AA151RpEdsNufCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f8952a7b8a4dbf45da7ee7024ff39f08d2bb9badb2a808af4629886b18748766","last_reissued_at":"2026-08-06T01:47:43.963239Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:47:43.963239Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Unleashing the Potential of Vision-Language Models for Generalizable AI-Generated Image Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hao Tan, Jun Wan, Weihan Cai, Xinping Gao, Zichang Tan","submitted_at":"2026-08-05T15:03:44Z","abstract_excerpt":"Recent work has shown that a simple linear probe on frozen representations from modern vision foundation models (VFMs) can achieve state-of-the-art AIGI detection performance, substantially outperforming specialized detectors in challenging in-the-wild scenarios. This finding has established DINOv3 as the dominant foundation-model baseline for subsequent improvements. However, we find that the vision-language model Perception Encoder (PE) holds greater potential for AIGI detection, because its language-aligned representation preserves high-level provenance semantics. Specifically, PE exhibits "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04935","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/2608.04935/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":"2608.04935","created_at":"2026-08-06T01:47:43.965043+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04935v1","created_at":"2026-08-06T01:47:43.965043+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04935","created_at":"2026-08-06T01:47:43.965043+00:00"},{"alias_kind":"pith_short_12","alias_value":"7CKSU64KJW7U","created_at":"2026-08-06T01:47:43.965043+00:00"},{"alias_kind":"pith_short_16","alias_value":"7CKSU64KJW7ULWT6","created_at":"2026-08-06T01:47:43.965043+00:00"},{"alias_kind":"pith_short_8","alias_value":"7CKSU64K","created_at":"2026-08-06T01:47:43.965043+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/7CKSU64KJW7ULWT644BE7447BD","json":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD.json","graph_json":"https://pith.science/api/pith-number/7CKSU64KJW7ULWT644BE7447BD/graph.json","events_json":"https://pith.science/api/pith-number/7CKSU64KJW7ULWT644BE7447BD/events.json","paper":"https://pith.science/paper/7CKSU64K"},"agent_actions":{"view_html":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD","download_json":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD.json","view_paper":"https://pith.science/paper/7CKSU64K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04935&json=true","fetch_graph":"https://pith.science/api/pith-number/7CKSU64KJW7ULWT644BE7447BD/graph.json","fetch_events":"https://pith.science/api/pith-number/7CKSU64KJW7ULWT644BE7447BD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD/action/storage_attestation","attest_author":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD/action/author_attestation","sign_citation":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD/action/citation_signature","submit_replication":"https://pith.science/pith/7CKSU64KJW7ULWT644BE7447BD/action/replication_record"}},"created_at":"2026-08-06T01:47:43.965043+00:00","updated_at":"2026-08-06T01:47:43.965043+00:00"}