{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:ZWGRVIQGPI5HTYDA54QOJSQFNH","short_pith_number":"pith:ZWGRVIQG","schema_version":"1.0","canonical_sha256":"cd8d1aa2067a3a79e060ef20e4ca0569f1aef5a20ab05f6f1555ad70926796de","source":{"kind":"arxiv","id":"2207.00787","version":3},"attestation_state":"computed","paper":{"title":"Object Representations as Fixed Points: Training Iterative Refinement Algorithms with Implicit Differentiation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.NE"],"primary_cat":"cs.LG","authors_text":"Michael Chang, Sergey Levine, Thomas L. Griffiths","submitted_at":"2022-07-02T10:00:35Z","abstract_excerpt":"Iterative refinement -- start with a random guess, then iteratively improve the guess -- is a useful paradigm for representation learning because it offers a way to break symmetries among equally plausible explanations for the data. This property enables the application of such methods to infer representations of sets of entities, such as objects in physical scenes, structurally resembling clustering algorithms in latent space. However, most prior works differentiate through the unrolled refinement process, which can make optimization challenging. We observe that such methods can be made diffe"},"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":"2207.00787","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-07-02T10:00:35Z","cross_cats_sorted":["cs.AI","cs.CV","cs.NE"],"title_canon_sha256":"3101c620de857895320f1af4045fcdb4293b756396328522bd60f1143b915c4a","abstract_canon_sha256":"0194a2286856a756711b48f2bb9755ee22eca8060236d306442b2b101bb35720"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:29:28.271356Z","signature_b64":"ME9FE4TaAaiUZepGQoJsvGRThdLJGRxN9DIm/092hkDpBEFSvdoq/lK0uqCbM4qGnFmc7R1UG6r59g4A/u7PCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cd8d1aa2067a3a79e060ef20e4ca0569f1aef5a20ab05f6f1555ad70926796de","last_reissued_at":"2026-07-05T05:29:28.270866Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:29:28.270866Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Object Representations as Fixed Points: Training Iterative Refinement Algorithms with Implicit Differentiation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","cs.NE"],"primary_cat":"cs.LG","authors_text":"Michael Chang, Sergey Levine, Thomas L. Griffiths","submitted_at":"2022-07-02T10:00:35Z","abstract_excerpt":"Iterative refinement -- start with a random guess, then iteratively improve the guess -- is a useful paradigm for representation learning because it offers a way to break symmetries among equally plausible explanations for the data. This property enables the application of such methods to infer representations of sets of entities, such as objects in physical scenes, structurally resembling clustering algorithms in latent space. However, most prior works differentiate through the unrolled refinement process, which can make optimization challenging. We observe that such methods can be made diffe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.00787","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/2207.00787/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":"2207.00787","created_at":"2026-07-05T05:29:28.270924+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.00787v3","created_at":"2026-07-05T05:29:28.270924+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.00787","created_at":"2026-07-05T05:29:28.270924+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZWGRVIQGPI5H","created_at":"2026-07-05T05:29:28.270924+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZWGRVIQGPI5HTYDA","created_at":"2026-07-05T05:29:28.270924+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZWGRVIQG","created_at":"2026-07-05T05:29:28.270924+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.07806","citing_title":"Identifiable Object Representations under Spatial Ambiguities","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH","json":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH.json","graph_json":"https://pith.science/api/pith-number/ZWGRVIQGPI5HTYDA54QOJSQFNH/graph.json","events_json":"https://pith.science/api/pith-number/ZWGRVIQGPI5HTYDA54QOJSQFNH/events.json","paper":"https://pith.science/paper/ZWGRVIQG"},"agent_actions":{"view_html":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH","download_json":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH.json","view_paper":"https://pith.science/paper/ZWGRVIQG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.00787&json=true","fetch_graph":"https://pith.science/api/pith-number/ZWGRVIQGPI5HTYDA54QOJSQFNH/graph.json","fetch_events":"https://pith.science/api/pith-number/ZWGRVIQGPI5HTYDA54QOJSQFNH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH/action/storage_attestation","attest_author":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH/action/author_attestation","sign_citation":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH/action/citation_signature","submit_replication":"https://pith.science/pith/ZWGRVIQGPI5HTYDA54QOJSQFNH/action/replication_record"}},"created_at":"2026-07-05T05:29:28.270924+00:00","updated_at":"2026-07-05T05:29:28.270924+00:00"}