{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UUVPTLZWIEMPOXYCIJO27RK3PT","short_pith_number":"pith:UUVPTLZW","schema_version":"1.0","canonical_sha256":"a52af9af364118f75f02425dafc55b7cf9d0b8d36f33db36a628d0a867c43200","source":{"kind":"arxiv","id":"2507.07985","version":2},"attestation_state":"computed","paper":{"title":"Common Data Properties Limit Object-Attribute Binding in CLIP","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bijay Gurung, David T. Hoffmann, Thomas Brox","submitted_at":"2025-07-10T17:57:31Z","abstract_excerpt":"Contrastive vision-language models like CLIP are used for a large variety of applications, such as zero-shot classification or as vision encoder for multi-modal models. Despite their popularity, their representations show major limitations. For instance, CLIP models learn bag-of-words representations and, as a consequence, fail to distinguish whether an image is of ``a yellow submarine and a blue bus'' or ``a blue submarine and a yellow bus''. Previous attempts to fix this issue added hard negatives during training or modified the architecture, but failed to resolve the problem in its entirety"},"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":"2507.07985","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-07-10T17:57:31Z","cross_cats_sorted":[],"title_canon_sha256":"df94725869d320d6b04b9e4051b11e63653faab03dadee98b8514d1db8ff0017","abstract_canon_sha256":"53f7aa0cb9ea248e485875b3a0f8428fcd5557628576c1b155936109f53168fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:53:48.525517Z","signature_b64":"H/V9+0bdRqShwda+b+aRlqoH2BlwzHKtgM0cyAvgEQ/UUptXldBHGEGYSClXV+TKlVe1cp2sBZjuf3n9ney9Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a52af9af364118f75f02425dafc55b7cf9d0b8d36f33db36a628d0a867c43200","last_reissued_at":"2026-07-05T11:53:48.525028Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:53:48.525028Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Common Data Properties Limit Object-Attribute Binding in CLIP","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bijay Gurung, David T. Hoffmann, Thomas Brox","submitted_at":"2025-07-10T17:57:31Z","abstract_excerpt":"Contrastive vision-language models like CLIP are used for a large variety of applications, such as zero-shot classification or as vision encoder for multi-modal models. Despite their popularity, their representations show major limitations. For instance, CLIP models learn bag-of-words representations and, as a consequence, fail to distinguish whether an image is of ``a yellow submarine and a blue bus'' or ``a blue submarine and a yellow bus''. Previous attempts to fix this issue added hard negatives during training or modified the architecture, but failed to resolve the problem in its entirety"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.07985","kind":"arxiv","version":2},"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/2507.07985/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":"2507.07985","created_at":"2026-07-05T11:53:48.525084+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.07985v2","created_at":"2026-07-05T11:53:48.525084+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.07985","created_at":"2026-07-05T11:53:48.525084+00:00"},{"alias_kind":"pith_short_12","alias_value":"UUVPTLZWIEMP","created_at":"2026-07-05T11:53:48.525084+00:00"},{"alias_kind":"pith_short_16","alias_value":"UUVPTLZWIEMPOXYC","created_at":"2026-07-05T11:53:48.525084+00:00"},{"alias_kind":"pith_short_8","alias_value":"UUVPTLZW","created_at":"2026-07-05T11:53:48.525084+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.20322","citing_title":"Disentangling Latent Embeddings with Sparse Linear Concept Subspaces (SLiCS)","ref_index":17,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT","json":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT.json","graph_json":"https://pith.science/api/pith-number/UUVPTLZWIEMPOXYCIJO27RK3PT/graph.json","events_json":"https://pith.science/api/pith-number/UUVPTLZWIEMPOXYCIJO27RK3PT/events.json","paper":"https://pith.science/paper/UUVPTLZW"},"agent_actions":{"view_html":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT","download_json":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT.json","view_paper":"https://pith.science/paper/UUVPTLZW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.07985&json=true","fetch_graph":"https://pith.science/api/pith-number/UUVPTLZWIEMPOXYCIJO27RK3PT/graph.json","fetch_events":"https://pith.science/api/pith-number/UUVPTLZWIEMPOXYCIJO27RK3PT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT/action/storage_attestation","attest_author":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT/action/author_attestation","sign_citation":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT/action/citation_signature","submit_replication":"https://pith.science/pith/UUVPTLZWIEMPOXYCIJO27RK3PT/action/replication_record"}},"created_at":"2026-07-05T11:53:48.525084+00:00","updated_at":"2026-07-05T11:53:48.525084+00:00"}