{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:5K77QJ3EY5DIV46ZOH5PXLJOUA","short_pith_number":"pith:5K77QJ3E","schema_version":"1.0","canonical_sha256":"eabff82764c7468af3d971fafbad2ea0341c32869e2064b3b68a0c043cecb1de","source":{"kind":"arxiv","id":"2205.15021","version":1},"attestation_state":"computed","paper":{"title":"Agnostic Physics-Driven Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Benjamin Scellier, Siddhartha Mishra, Yann Ollivier, Yoshua Bengio","submitted_at":"2022-05-30T12:02:53Z","abstract_excerpt":"This work establishes that a physical system can perform statistical learning without gradient computations, via an Agnostic Equilibrium Propagation (Aeqprop) procedure that combines energy minimization, homeostatic control, and nudging towards the correct response. In Aeqprop, the specifics of the system do not have to be known: the procedure is based only on external manipulations, and produces a stochastic gradient descent without explicit gradient computations. Thanks to nudging, the system performs a true, order-one gradient step for each training sample, in contrast with order-zero metho"},"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":"2205.15021","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-05-30T12:02:53Z","cross_cats_sorted":[],"title_canon_sha256":"35ce9c4068defe2e715ab384ce1582174deee8a5a30141cdb8f3834e9a38ebae","abstract_canon_sha256":"4cc61da4dd59463b88194be7acd5a53e491dd9f53e67e629be2dee781ea2c6f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:27:23.544967Z","signature_b64":"T7MervswUSU5KIUNoxAVvxKFIjc5BzaK8agMLm+AeisS3qfb7R4zFDKbaIgLq4HY84bbQmb0g7WMN7LnXXQSBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eabff82764c7468af3d971fafbad2ea0341c32869e2064b3b68a0c043cecb1de","last_reissued_at":"2026-07-05T04:27:23.544508Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:27:23.544508Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Agnostic Physics-Driven Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Benjamin Scellier, Siddhartha Mishra, Yann Ollivier, Yoshua Bengio","submitted_at":"2022-05-30T12:02:53Z","abstract_excerpt":"This work establishes that a physical system can perform statistical learning without gradient computations, via an Agnostic Equilibrium Propagation (Aeqprop) procedure that combines energy minimization, homeostatic control, and nudging towards the correct response. In Aeqprop, the specifics of the system do not have to be known: the procedure is based only on external manipulations, and produces a stochastic gradient descent without explicit gradient computations. Thanks to nudging, the system performs a true, order-one gradient step for each training sample, in contrast with order-zero metho"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.15021","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/2205.15021/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":"2205.15021","created_at":"2026-07-05T04:27:23.544560+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.15021v1","created_at":"2026-07-05T04:27:23.544560+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.15021","created_at":"2026-07-05T04:27:23.544560+00:00"},{"alias_kind":"pith_short_12","alias_value":"5K77QJ3EY5DI","created_at":"2026-07-05T04:27:23.544560+00:00"},{"alias_kind":"pith_short_16","alias_value":"5K77QJ3EY5DIV46Z","created_at":"2026-07-05T04:27:23.544560+00:00"},{"alias_kind":"pith_short_8","alias_value":"5K77QJ3E","created_at":"2026-07-05T04:27:23.544560+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.13454","citing_title":"Optical Implementation of Equilibrium Propagation Using Spatial Photonic Ising Machines","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.09112","citing_title":"Hybridizing Equilibrium Propagation with Ising Machines for Efficient Energy-Based Learning","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03584","citing_title":"Training a Predictive Coding Network on ImageNet using Equilibrium Propagation","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA","json":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA.json","graph_json":"https://pith.science/api/pith-number/5K77QJ3EY5DIV46ZOH5PXLJOUA/graph.json","events_json":"https://pith.science/api/pith-number/5K77QJ3EY5DIV46ZOH5PXLJOUA/events.json","paper":"https://pith.science/paper/5K77QJ3E"},"agent_actions":{"view_html":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA","download_json":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA.json","view_paper":"https://pith.science/paper/5K77QJ3E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.15021&json=true","fetch_graph":"https://pith.science/api/pith-number/5K77QJ3EY5DIV46ZOH5PXLJOUA/graph.json","fetch_events":"https://pith.science/api/pith-number/5K77QJ3EY5DIV46ZOH5PXLJOUA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA/action/storage_attestation","attest_author":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA/action/author_attestation","sign_citation":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA/action/citation_signature","submit_replication":"https://pith.science/pith/5K77QJ3EY5DIV46ZOH5PXLJOUA/action/replication_record"}},"created_at":"2026-07-05T04:27:23.544560+00:00","updated_at":"2026-07-05T04:27:23.544560+00:00"}