{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:LLHKVKW37ZWWO35IDRARKZFV3T","short_pith_number":"pith:LLHKVKW3","schema_version":"1.0","canonical_sha256":"5aceaaaadbfe6d676fa81c411564b5dcdc7b09210a06b016330dfc3c7645e139","source":{"kind":"arxiv","id":"1912.05783","version":2},"attestation_state":"computed","paper":{"title":"CLOSURE: Assessing Systematic Generalization of CLEVR Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Aaron Courville, Dzmitry Bahdanau, Harm de Vries, Philippe Beaudoin, Shikhar Murty, Timothy J. O'Donnell, Yoshua Bengio","submitted_at":"2019-12-12T05:56:53Z","abstract_excerpt":"The CLEVR dataset of natural-looking questions about 3D-rendered scenes has recently received much attention from the research community. A number of models have been proposed for this task, many of which achieved very high accuracies of around 97-99%. In this work, we study how systematic the generalization of such models is, that is to which extent they are capable of handling novel combinations of known linguistic constructs. To this end, we test models' understanding of referring expressions based on matching object properties (such as e.g. \"another cube that is the same size as the brown "},"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":"1912.05783","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2019-12-12T05:56:53Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"22d53f3aa6f80ac25f6b9ee67f006f9d28095c433f1b1f75e2857c16e664f9a1","abstract_canon_sha256":"26e450dd33c927d4bacc6a20db6dcfde0acd0080d9963b2f241bf86690e664c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:43:38.585708Z","signature_b64":"fuznFTBfuTbpX78l4vzziaWrtO58pWdULHANFh//sWtwVZOgFpPAHJYgEPaj1vsrW++3maB9PRuqxgZ9zJuFCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5aceaaaadbfe6d676fa81c411564b5dcdc7b09210a06b016330dfc3c7645e139","last_reissued_at":"2026-07-05T01:43:38.585266Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:43:38.585266Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CLOSURE: Assessing Systematic Generalization of CLEVR Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Aaron Courville, Dzmitry Bahdanau, Harm de Vries, Philippe Beaudoin, Shikhar Murty, Timothy J. O'Donnell, Yoshua Bengio","submitted_at":"2019-12-12T05:56:53Z","abstract_excerpt":"The CLEVR dataset of natural-looking questions about 3D-rendered scenes has recently received much attention from the research community. A number of models have been proposed for this task, many of which achieved very high accuracies of around 97-99%. In this work, we study how systematic the generalization of such models is, that is to which extent they are capable of handling novel combinations of known linguistic constructs. To this end, we test models' understanding of referring expressions based on matching object properties (such as e.g. \"another cube that is the same size as the brown "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.05783","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/1912.05783/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":"1912.05783","created_at":"2026-07-05T01:43:38.585335+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.05783v2","created_at":"2026-07-05T01:43:38.585335+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.05783","created_at":"2026-07-05T01:43:38.585335+00:00"},{"alias_kind":"pith_short_12","alias_value":"LLHKVKW37ZWW","created_at":"2026-07-05T01:43:38.585335+00:00"},{"alias_kind":"pith_short_16","alias_value":"LLHKVKW37ZWWO35I","created_at":"2026-07-05T01:43:38.585335+00:00"},{"alias_kind":"pith_short_8","alias_value":"LLHKVKW3","created_at":"2026-07-05T01:43:38.585335+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.02168","citing_title":"Disentanglement-Based Equivariant Learning for Compositional VQA","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2509.19207","citing_title":"Long Story Short: Disentangling Compositionality and Long-Caption Understanding in Contrastive VLMs","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11680","citing_title":"ShapeCodeBench: A Renewable Benchmark for Perception-to-Program Reconstruction of Synthetic Shape Scenes","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T","json":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T.json","graph_json":"https://pith.science/api/pith-number/LLHKVKW37ZWWO35IDRARKZFV3T/graph.json","events_json":"https://pith.science/api/pith-number/LLHKVKW37ZWWO35IDRARKZFV3T/events.json","paper":"https://pith.science/paper/LLHKVKW3"},"agent_actions":{"view_html":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T","download_json":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T.json","view_paper":"https://pith.science/paper/LLHKVKW3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.05783&json=true","fetch_graph":"https://pith.science/api/pith-number/LLHKVKW37ZWWO35IDRARKZFV3T/graph.json","fetch_events":"https://pith.science/api/pith-number/LLHKVKW37ZWWO35IDRARKZFV3T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T/action/storage_attestation","attest_author":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T/action/author_attestation","sign_citation":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T/action/citation_signature","submit_replication":"https://pith.science/pith/LLHKVKW37ZWWO35IDRARKZFV3T/action/replication_record"}},"created_at":"2026-07-05T01:43:38.585335+00:00","updated_at":"2026-07-05T01:43:38.585335+00:00"}