{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:3GSRMQEW72QQFZK7OABUMGM5ZG","short_pith_number":"pith:3GSRMQEW","schema_version":"1.0","canonical_sha256":"d9a5164096fea102e55f700346199dc98b19a28d42b814cd58057f6e42e65a4b","source":{"kind":"arxiv","id":"2107.09106","version":1},"attestation_state":"computed","paper":{"title":"Separating Skills and Concepts for Novel Visual Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Heng Ji, Hui Wu, Kate Saenko, Rogerio Feris, Spencer Whitehead","submitted_at":"2021-07-19T18:55:10Z","abstract_excerpt":"Generalization to out-of-distribution data has been a problem for Visual Question Answering (VQA) models. To measure generalization to novel questions, we propose to separate them into \"skills\" and \"concepts\". \"Skills\" are visual tasks, such as counting or attribute recognition, and are applied to \"concepts\" mentioned in the question, such as objects and people. VQA methods should be able to compose skills and concepts in novel ways, regardless of whether the specific composition has been seen in training, yet we demonstrate that existing models have much to improve upon towards handling new 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":"2107.09106","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-07-19T18:55:10Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"916c52d5e083f2c91926c8119c95a6f06c5d69b536f1a611b198599a7a24d5f9","abstract_canon_sha256":"e312498c46b2ee4ad977822276202b550831048be73e8fba47af0611e92970ef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:59:26.066353Z","signature_b64":"VRFUo9crEJZZechnfSWvX6uRt7KfnXgyXrah5Dl9RVg3QNJVAMlb4A6tQ3ApgQNpbVPr/EksXkLoOfwfE+LlDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d9a5164096fea102e55f700346199dc98b19a28d42b814cd58057f6e42e65a4b","last_reissued_at":"2026-07-05T02:59:26.065877Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:59:26.065877Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Separating Skills and Concepts for Novel Visual Question Answering","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Heng Ji, Hui Wu, Kate Saenko, Rogerio Feris, Spencer Whitehead","submitted_at":"2021-07-19T18:55:10Z","abstract_excerpt":"Generalization to out-of-distribution data has been a problem for Visual Question Answering (VQA) models. To measure generalization to novel questions, we propose to separate them into \"skills\" and \"concepts\". \"Skills\" are visual tasks, such as counting or attribute recognition, and are applied to \"concepts\" mentioned in the question, such as objects and people. VQA methods should be able to compose skills and concepts in novel ways, regardless of whether the specific composition has been seen in training, yet we demonstrate that existing models have much to improve upon towards handling new c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2107.09106","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/2107.09106/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":"2107.09106","created_at":"2026-07-05T02:59:26.065933+00:00"},{"alias_kind":"arxiv_version","alias_value":"2107.09106v1","created_at":"2026-07-05T02:59:26.065933+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2107.09106","created_at":"2026-07-05T02:59:26.065933+00:00"},{"alias_kind":"pith_short_12","alias_value":"3GSRMQEW72QQ","created_at":"2026-07-05T02:59:26.065933+00:00"},{"alias_kind":"pith_short_16","alias_value":"3GSRMQEW72QQFZK7","created_at":"2026-07-05T02:59:26.065933+00:00"},{"alias_kind":"pith_short_8","alias_value":"3GSRMQEW","created_at":"2026-07-05T02:59:26.065933+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.10339","citing_title":"Concepts or Skills? Rethinking Instruction Selection for Multi-modal Models","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG","json":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG.json","graph_json":"https://pith.science/api/pith-number/3GSRMQEW72QQFZK7OABUMGM5ZG/graph.json","events_json":"https://pith.science/api/pith-number/3GSRMQEW72QQFZK7OABUMGM5ZG/events.json","paper":"https://pith.science/paper/3GSRMQEW"},"agent_actions":{"view_html":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG","download_json":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG.json","view_paper":"https://pith.science/paper/3GSRMQEW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2107.09106&json=true","fetch_graph":"https://pith.science/api/pith-number/3GSRMQEW72QQFZK7OABUMGM5ZG/graph.json","fetch_events":"https://pith.science/api/pith-number/3GSRMQEW72QQFZK7OABUMGM5ZG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG/action/storage_attestation","attest_author":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG/action/author_attestation","sign_citation":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG/action/citation_signature","submit_replication":"https://pith.science/pith/3GSRMQEW72QQFZK7OABUMGM5ZG/action/replication_record"}},"created_at":"2026-07-05T02:59:26.065933+00:00","updated_at":"2026-07-05T02:59:26.065933+00:00"}