{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:4Q2BUKIB7VKUYVI27DUCJEYOTK","short_pith_number":"pith:4Q2BUKIB","schema_version":"1.0","canonical_sha256":"e4341a2901fd554c551af8e824930e9ab198902593a144078f8252f6cddf63fb","source":{"kind":"arxiv","id":"2104.08945","version":1},"attestation_state":"computed","paper":{"title":"Data-Efficient Language-Supervised Zero-Shot Learning with Self-Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bichen Wu, Joseph E. Gonzalez, Peizhao Zhang, Peter Vajda, Ruizhe Cheng","submitted_at":"2021-04-18T19:55:31Z","abstract_excerpt":"Traditional computer vision models are trained to predict a fixed set of predefined categories. Recently, natural language has been shown to be a broader and richer source of supervision that provides finer descriptions to visual concepts than supervised \"gold\" labels. Previous works, such as CLIP, use a simple pretraining task of predicting the pairings between images and text captions. CLIP, however, is data hungry and requires more than 400M image text pairs for training. We propose a data-efficient contrastive distillation method that uses soft labels to learn from noisy image-text pairs. "},"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":"2104.08945","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2021-04-18T19:55:31Z","cross_cats_sorted":[],"title_canon_sha256":"ce818cbda3decb75c35fe26e5422c9c4e0f74b836c2335201eaa6a92ca4a95ab","abstract_canon_sha256":"b9c3238794487842e1599f5f3595c9314919df767b53240485bca87d4481d614"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:32:57.093960Z","signature_b64":"ZNYiHvu7XNQEwe2NnLmugRTHhh7Sgz9wVu9rYcepYnbj2VNd17ZuFJHNWI74Z+FCIpozHP1LR0OG5I50SP8uDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4341a2901fd554c551af8e824930e9ab198902593a144078f8252f6cddf63fb","last_reissued_at":"2026-07-05T02:32:57.093476Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:32:57.093476Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Data-Efficient Language-Supervised Zero-Shot Learning with Self-Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bichen Wu, Joseph E. Gonzalez, Peizhao Zhang, Peter Vajda, Ruizhe Cheng","submitted_at":"2021-04-18T19:55:31Z","abstract_excerpt":"Traditional computer vision models are trained to predict a fixed set of predefined categories. Recently, natural language has been shown to be a broader and richer source of supervision that provides finer descriptions to visual concepts than supervised \"gold\" labels. Previous works, such as CLIP, use a simple pretraining task of predicting the pairings between images and text captions. CLIP, however, is data hungry and requires more than 400M image text pairs for training. We propose a data-efficient contrastive distillation method that uses soft labels to learn from noisy image-text pairs. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.08945","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/2104.08945/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":"2104.08945","created_at":"2026-07-05T02:32:57.093538+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.08945v1","created_at":"2026-07-05T02:32:57.093538+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.08945","created_at":"2026-07-05T02:32:57.093538+00:00"},{"alias_kind":"pith_short_12","alias_value":"4Q2BUKIB7VKU","created_at":"2026-07-05T02:32:57.093538+00:00"},{"alias_kind":"pith_short_16","alias_value":"4Q2BUKIB7VKUYVI2","created_at":"2026-07-05T02:32:57.093538+00:00"},{"alias_kind":"pith_short_8","alias_value":"4Q2BUKIB","created_at":"2026-07-05T02:32:57.093538+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/4Q2BUKIB7VKUYVI27DUCJEYOTK","json":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK.json","graph_json":"https://pith.science/api/pith-number/4Q2BUKIB7VKUYVI27DUCJEYOTK/graph.json","events_json":"https://pith.science/api/pith-number/4Q2BUKIB7VKUYVI27DUCJEYOTK/events.json","paper":"https://pith.science/paper/4Q2BUKIB"},"agent_actions":{"view_html":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK","download_json":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK.json","view_paper":"https://pith.science/paper/4Q2BUKIB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.08945&json=true","fetch_graph":"https://pith.science/api/pith-number/4Q2BUKIB7VKUYVI27DUCJEYOTK/graph.json","fetch_events":"https://pith.science/api/pith-number/4Q2BUKIB7VKUYVI27DUCJEYOTK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK/action/storage_attestation","attest_author":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK/action/author_attestation","sign_citation":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK/action/citation_signature","submit_replication":"https://pith.science/pith/4Q2BUKIB7VKUYVI27DUCJEYOTK/action/replication_record"}},"created_at":"2026-07-05T02:32:57.093538+00:00","updated_at":"2026-07-05T02:32:57.093538+00:00"}