{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LRQSKRK72USEY3NKAIAKY4O3BS","short_pith_number":"pith:LRQSKRK7","schema_version":"1.0","canonical_sha256":"5c6125455fd5244c6daa0200ac71db0c8b05f5d6da08bf5b77155650472bd353","source":{"kind":"arxiv","id":"2312.07661","version":3},"attestation_state":"computed","paper":{"title":"CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Philip Torr, Runjia Li, Shuyang Sun, Siyang Li, Xiuye Gu","submitted_at":"2023-12-12T19:00:04Z","abstract_excerpt":"Existing open-vocabulary image segmentation methods require a fine-tuning step on mask labels and/or image-text datasets. Mask labels are labor-intensive, which limits the number of categories in segmentation datasets. Consequently, the vocabulary capacity of pre-trained VLMs is severely reduced after fine-tuning. However, without fine-tuning, VLMs trained under weak image-text supervision tend to make suboptimal mask predictions. To alleviate these issues, we introduce a novel recurrent framework that progressively filters out irrelevant texts and enhances mask quality without training effort"},"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":"2312.07661","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-12-12T19:00:04Z","cross_cats_sorted":["cs.CL","cs.LG","cs.MM"],"title_canon_sha256":"89edd37a8a45c93e463ed4c0376a0ecd9fdaa0e7cf1acee28737940a4331a34f","abstract_canon_sha256":"853cce57460b07cf22538e8c81ae3b74376e6594f9fb670b2fa901e64438f63a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:16:24.209329Z","signature_b64":"Fkp8iguRTpqrDy5pDlzd8foKUizKxSBZfjo5Oi3OEc36ytss3SRgSYTXAxY8HscZJnEdHbsr8cFIXzz8f3erCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5c6125455fd5244c6daa0200ac71db0c8b05f5d6da08bf5b77155650472bd353","last_reissued_at":"2026-07-05T08:16:24.208877Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:16:24.208877Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.MM"],"primary_cat":"cs.CV","authors_text":"Philip Torr, Runjia Li, Shuyang Sun, Siyang Li, Xiuye Gu","submitted_at":"2023-12-12T19:00:04Z","abstract_excerpt":"Existing open-vocabulary image segmentation methods require a fine-tuning step on mask labels and/or image-text datasets. Mask labels are labor-intensive, which limits the number of categories in segmentation datasets. Consequently, the vocabulary capacity of pre-trained VLMs is severely reduced after fine-tuning. However, without fine-tuning, VLMs trained under weak image-text supervision tend to make suboptimal mask predictions. To alleviate these issues, we introduce a novel recurrent framework that progressively filters out irrelevant texts and enhances mask quality without training effort"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.07661","kind":"arxiv","version":3},"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/2312.07661/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":"2312.07661","created_at":"2026-07-05T08:16:24.208938+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.07661v3","created_at":"2026-07-05T08:16:24.208938+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.07661","created_at":"2026-07-05T08:16:24.208938+00:00"},{"alias_kind":"pith_short_12","alias_value":"LRQSKRK72USE","created_at":"2026-07-05T08:16:24.208938+00:00"},{"alias_kind":"pith_short_16","alias_value":"LRQSKRK72USEY3NK","created_at":"2026-07-05T08:16:24.208938+00:00"},{"alias_kind":"pith_short_8","alias_value":"LRQSKRK7","created_at":"2026-07-05T08:16:24.208938+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.04320","citing_title":"ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features","ref_index":46,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS","json":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS.json","graph_json":"https://pith.science/api/pith-number/LRQSKRK72USEY3NKAIAKY4O3BS/graph.json","events_json":"https://pith.science/api/pith-number/LRQSKRK72USEY3NKAIAKY4O3BS/events.json","paper":"https://pith.science/paper/LRQSKRK7"},"agent_actions":{"view_html":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS","download_json":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS.json","view_paper":"https://pith.science/paper/LRQSKRK7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.07661&json=true","fetch_graph":"https://pith.science/api/pith-number/LRQSKRK72USEY3NKAIAKY4O3BS/graph.json","fetch_events":"https://pith.science/api/pith-number/LRQSKRK72USEY3NKAIAKY4O3BS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS/action/storage_attestation","attest_author":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS/action/author_attestation","sign_citation":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS/action/citation_signature","submit_replication":"https://pith.science/pith/LRQSKRK72USEY3NKAIAKY4O3BS/action/replication_record"}},"created_at":"2026-07-05T08:16:24.208938+00:00","updated_at":"2026-07-05T08:16:24.208938+00:00"}