{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DQWMDIF3BVNJEFGWB2U2HKHQAL","short_pith_number":"pith:DQWMDIF3","schema_version":"1.0","canonical_sha256":"1c2cc1a0bb0d5a9214d60ea9a3a8f002f9c271e3a1d6a17a7b5d1c66a2343ec4","source":{"kind":"arxiv","id":"2407.12442","version":1},"attestation_state":"computed","paper":{"title":"ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chaofeng Chen, Litong Feng, Mengcheng Lan, Wayne Zhang, Xinjiang Wang, Yiping Ke","submitted_at":"2024-07-17T09:52:20Z","abstract_excerpt":"Despite the success of large-scale pretrained Vision-Language Models (VLMs) especially CLIP in various open-vocabulary tasks, their application to semantic segmentation remains challenging, producing noisy segmentation maps with mis-segmented regions. In this paper, we carefully re-investigate the architecture of CLIP, and identify residual connections as the primary source of noise that degrades segmentation quality. With a comparative analysis of statistical properties in the residual connection and the attention output across different pretrained models, we discover that CLIP's image-text c"},"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":"2407.12442","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-07-17T09:52:20Z","cross_cats_sorted":[],"title_canon_sha256":"65ee95b69c18dfc261223f628b60065ab4b51e1043be553151d4a066deda283e","abstract_canon_sha256":"69734dea8b957592e1aadd6a40238910fbdb45dca6e34ef042a3ef13cd46436a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:45:04.662811Z","signature_b64":"rrG5nzlbL+mGT/FdXv5J3HFkD6jFyj6jhghAuqo9EG5C3PSxm91r/O9apxqDlvclvw+hKddV7YI3Rip0ms3ZAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1c2cc1a0bb0d5a9214d60ea9a3a8f002f9c271e3a1d6a17a7b5d1c66a2343ec4","last_reissued_at":"2026-07-05T08:45:04.662401Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:45:04.662401Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chaofeng Chen, Litong Feng, Mengcheng Lan, Wayne Zhang, Xinjiang Wang, Yiping Ke","submitted_at":"2024-07-17T09:52:20Z","abstract_excerpt":"Despite the success of large-scale pretrained Vision-Language Models (VLMs) especially CLIP in various open-vocabulary tasks, their application to semantic segmentation remains challenging, producing noisy segmentation maps with mis-segmented regions. In this paper, we carefully re-investigate the architecture of CLIP, and identify residual connections as the primary source of noise that degrades segmentation quality. With a comparative analysis of statistical properties in the residual connection and the attention output across different pretrained models, we discover that CLIP's image-text c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.12442","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/2407.12442/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":"2407.12442","created_at":"2026-07-05T08:45:04.662457+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.12442v1","created_at":"2026-07-05T08:45:04.662457+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.12442","created_at":"2026-07-05T08:45:04.662457+00:00"},{"alias_kind":"pith_short_12","alias_value":"DQWMDIF3BVNJ","created_at":"2026-07-05T08:45:04.662457+00:00"},{"alias_kind":"pith_short_16","alias_value":"DQWMDIF3BVNJEFGW","created_at":"2026-07-05T08:45:04.662457+00:00"},{"alias_kind":"pith_short_8","alias_value":"DQWMDIF3","created_at":"2026-07-05T08:45:04.662457+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2408.06747","citing_title":"ReCLIP++: Learn to Rectify the Bias of CLIP for Unsupervised Semantic Segmentation","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2502.06818","citing_title":"Rethinking the Global Knowledge of CLIP in Training-Free Open-Vocabulary Semantic Segmentation","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2512.08730","citing_title":"SegEarth-OV3: Exploring SAM 3 for Open-Vocabulary Semantic Segmentation in Remote Sensing Images","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL","json":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL.json","graph_json":"https://pith.science/api/pith-number/DQWMDIF3BVNJEFGWB2U2HKHQAL/graph.json","events_json":"https://pith.science/api/pith-number/DQWMDIF3BVNJEFGWB2U2HKHQAL/events.json","paper":"https://pith.science/paper/DQWMDIF3"},"agent_actions":{"view_html":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL","download_json":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL.json","view_paper":"https://pith.science/paper/DQWMDIF3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.12442&json=true","fetch_graph":"https://pith.science/api/pith-number/DQWMDIF3BVNJEFGWB2U2HKHQAL/graph.json","fetch_events":"https://pith.science/api/pith-number/DQWMDIF3BVNJEFGWB2U2HKHQAL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL/action/storage_attestation","attest_author":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL/action/author_attestation","sign_citation":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL/action/citation_signature","submit_replication":"https://pith.science/pith/DQWMDIF3BVNJEFGWB2U2HKHQAL/action/replication_record"}},"created_at":"2026-07-05T08:45:04.662457+00:00","updated_at":"2026-07-05T08:45:04.662457+00:00"}