{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XDE6OBOGKS3IVB6IL5DFZO56FE","short_pith_number":"pith:XDE6OBOG","schema_version":"1.0","canonical_sha256":"b8c9e705c654b68a87c85f465cbbbe293efa0a1cb625afffb974efb3de49944a","source":{"kind":"arxiv","id":"2403.11998","version":2},"attestation_state":"computed","paper":{"title":"Learning Useful Representations of Recurrent Neural Network Weight Matrices","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Francesco Faccio, J\\\"urgen Schmidhuber, Vincent Herrmann","submitted_at":"2024-03-18T17:32:23Z","abstract_excerpt":"Recurrent Neural Networks (RNNs) are general-purpose parallel-sequential computers. The program of an RNN is its weight matrix. How to learn useful representations of RNN weights that facilitate RNN analysis as well as downstream tasks? While the mechanistic approach directly looks at some RNN's weights to predict its behavior, the functionalist approach analyzes its overall functionality-specifically, its input-output mapping. We consider several mechanistic approaches for RNN weights and adapt the permutation equivariant Deep Weight Space layer for RNNs. Our two novel functionalist approache"},"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":"2403.11998","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-18T17:32:23Z","cross_cats_sorted":[],"title_canon_sha256":"a6fa6ed27c5c6b8fd18981d776f743203d0b3c756fe0f4eb0f55d757f48b8e61","abstract_canon_sha256":"f2dbc16ff90d317a4ee91c26f2e9fc2ab8830d743a24885a2d6803588f71b78e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:55:18.605215Z","signature_b64":"XEx3bgoSazECpkP4EvNJAhqkxaZ277HQVfIqV9c/CI9zlUPEzdVAXpUaBxLXG7W7ItMcUphShFnDj8rHhkH5Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b8c9e705c654b68a87c85f465cbbbe293efa0a1cb625afffb974efb3de49944a","last_reissued_at":"2026-07-05T10:55:18.604712Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:55:18.604712Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Useful Representations of Recurrent Neural Network Weight Matrices","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Francesco Faccio, J\\\"urgen Schmidhuber, Vincent Herrmann","submitted_at":"2024-03-18T17:32:23Z","abstract_excerpt":"Recurrent Neural Networks (RNNs) are general-purpose parallel-sequential computers. The program of an RNN is its weight matrix. How to learn useful representations of RNN weights that facilitate RNN analysis as well as downstream tasks? While the mechanistic approach directly looks at some RNN's weights to predict its behavior, the functionalist approach analyzes its overall functionality-specifically, its input-output mapping. We consider several mechanistic approaches for RNN weights and adapt the permutation equivariant Deep Weight Space layer for RNNs. Our two novel functionalist approache"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.11998","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/2403.11998/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":"2403.11998","created_at":"2026-07-05T10:55:18.604766+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.11998v2","created_at":"2026-07-05T10:55:18.604766+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.11998","created_at":"2026-07-05T10:55:18.604766+00:00"},{"alias_kind":"pith_short_12","alias_value":"XDE6OBOGKS3I","created_at":"2026-07-05T10:55:18.604766+00:00"},{"alias_kind":"pith_short_16","alias_value":"XDE6OBOGKS3IVB6I","created_at":"2026-07-05T10:55:18.604766+00:00"},{"alias_kind":"pith_short_8","alias_value":"XDE6OBOG","created_at":"2026-07-05T10:55:18.604766+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.09619","citing_title":"Can this Model Also Recognize Dogs? Zero-Shot Model Search from Weights","ref_index":21,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE","json":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE.json","graph_json":"https://pith.science/api/pith-number/XDE6OBOGKS3IVB6IL5DFZO56FE/graph.json","events_json":"https://pith.science/api/pith-number/XDE6OBOGKS3IVB6IL5DFZO56FE/events.json","paper":"https://pith.science/paper/XDE6OBOG"},"agent_actions":{"view_html":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE","download_json":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE.json","view_paper":"https://pith.science/paper/XDE6OBOG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.11998&json=true","fetch_graph":"https://pith.science/api/pith-number/XDE6OBOGKS3IVB6IL5DFZO56FE/graph.json","fetch_events":"https://pith.science/api/pith-number/XDE6OBOGKS3IVB6IL5DFZO56FE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE/action/storage_attestation","attest_author":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE/action/author_attestation","sign_citation":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE/action/citation_signature","submit_replication":"https://pith.science/pith/XDE6OBOGKS3IVB6IL5DFZO56FE/action/replication_record"}},"created_at":"2026-07-05T10:55:18.604766+00:00","updated_at":"2026-07-05T10:55:18.604766+00:00"}