{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3NLPO4RKAX4TI3A2JI6ZAW5TAL","short_pith_number":"pith:3NLPO4RK","schema_version":"1.0","canonical_sha256":"db56f7722a05f9346c1a4a3d905bb302e40685ee8fe27adee114cff559999056","source":{"kind":"arxiv","id":"2504.11457","version":1},"attestation_state":"computed","paper":{"title":"Aligning Generative Denoising with Discriminative Objectives Unleashes Diffusion for Visual Perception","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Xin Xu, Yu-Xiong Wang, Ziqi Pang","submitted_at":"2025-04-15T17:59:54Z","abstract_excerpt":"With the success of image generation, generative diffusion models are increasingly adopted for discriminative tasks, as pixel generation provides a unified perception interface. However, directly repurposing the generative denoising process for discriminative objectives reveals critical gaps rarely addressed previously. Generative models tolerate intermediate sampling errors if the final distribution remains plausible, but discriminative tasks require rigorous accuracy throughout, as evidenced in challenging multi-modal tasks like referring image segmentation. Motivated by this gap, we analyze"},"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":"2504.11457","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-04-15T17:59:54Z","cross_cats_sorted":[],"title_canon_sha256":"07e50d51b9617fd10fc3e55830045eebbd21aec146955008953594fba98758ac","abstract_canon_sha256":"ae2f8c7d989ed3f832030a5756dee7882d32aee5e7eb51d1774aac85ba10a30c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:49:34.206950Z","signature_b64":"pRViH4X/5rcLw7y/BLwSX4trD5PJYGfr2sNU5ZRUEF1Enz9pIxzPzflikvRBza5S9tHmXPKZqH07LtyF+hDSAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db56f7722a05f9346c1a4a3d905bb302e40685ee8fe27adee114cff559999056","last_reissued_at":"2026-07-05T10:49:34.206479Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:49:34.206479Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Aligning Generative Denoising with Discriminative Objectives Unleashes Diffusion for Visual Perception","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Xin Xu, Yu-Xiong Wang, Ziqi Pang","submitted_at":"2025-04-15T17:59:54Z","abstract_excerpt":"With the success of image generation, generative diffusion models are increasingly adopted for discriminative tasks, as pixel generation provides a unified perception interface. However, directly repurposing the generative denoising process for discriminative objectives reveals critical gaps rarely addressed previously. Generative models tolerate intermediate sampling errors if the final distribution remains plausible, but discriminative tasks require rigorous accuracy throughout, as evidenced in challenging multi-modal tasks like referring image segmentation. Motivated by this gap, we analyze"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11457","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/2504.11457/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":"2504.11457","created_at":"2026-07-05T10:49:34.206538+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.11457v1","created_at":"2026-07-05T10:49:34.206538+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11457","created_at":"2026-07-05T10:49:34.206538+00:00"},{"alias_kind":"pith_short_12","alias_value":"3NLPO4RKAX4T","created_at":"2026-07-05T10:49:34.206538+00:00"},{"alias_kind":"pith_short_16","alias_value":"3NLPO4RKAX4TI3A2","created_at":"2026-07-05T10:49:34.206538+00:00"},{"alias_kind":"pith_short_8","alias_value":"3NLPO4RK","created_at":"2026-07-05T10:49:34.206538+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19743","citing_title":"EngiAI: A Multi-Agent Framework and Benchmark Suite for LLM-Driven Engineering Design","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19743","citing_title":"EngiAI: A Multi-Agent Framework and Benchmark Suite for LLM-Driven Engineering Design","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05077","citing_title":"FlowDIS: Language-Guided Dichotomous Image Segmentation with Flow Matching","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05077","citing_title":"FlowDIS: Language-Guided Dichotomous Image Segmentation with Flow Matching","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19234","citing_title":"Learning to Credit the Right Steps: Objective-aware Process Optimization for Visual Generation","ref_index":31,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL","json":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL.json","graph_json":"https://pith.science/api/pith-number/3NLPO4RKAX4TI3A2JI6ZAW5TAL/graph.json","events_json":"https://pith.science/api/pith-number/3NLPO4RKAX4TI3A2JI6ZAW5TAL/events.json","paper":"https://pith.science/paper/3NLPO4RK"},"agent_actions":{"view_html":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL","download_json":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL.json","view_paper":"https://pith.science/paper/3NLPO4RK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.11457&json=true","fetch_graph":"https://pith.science/api/pith-number/3NLPO4RKAX4TI3A2JI6ZAW5TAL/graph.json","fetch_events":"https://pith.science/api/pith-number/3NLPO4RKAX4TI3A2JI6ZAW5TAL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL/action/storage_attestation","attest_author":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL/action/author_attestation","sign_citation":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL/action/citation_signature","submit_replication":"https://pith.science/pith/3NLPO4RKAX4TI3A2JI6ZAW5TAL/action/replication_record"}},"created_at":"2026-07-05T10:49:34.206538+00:00","updated_at":"2026-07-05T10:49:34.206538+00:00"}