{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:ZGZGR2QUCFAM3SWNGYRF6Y7Q2G","short_pith_number":"pith:ZGZGR2QU","schema_version":"1.0","canonical_sha256":"c9b268ea141140cdcacd36225f63f0d185136462041ebd231b3c2ce18c067542","source":{"kind":"arxiv","id":"1611.09434","version":2},"attestation_state":"computed","paper":{"title":"Input Switched Affine Networks: An RNN Architecture Designed for Interpretability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.NE"],"primary_cat":"cs.AI","authors_text":"David Sussillo, Jakob N. Foerster, Jan Chorowski, Jascha Sohl-Dickstein, Justin Gilmer","submitted_at":"2016-11-28T23:48:41Z","abstract_excerpt":"There exist many problem domains where the interpretability of neural network models is essential for deployment. Here we introduce a recurrent architecture composed of input-switched affine transformations - in other words an RNN without any explicit nonlinearities, but with input-dependent recurrent weights. This simple form allows the RNN to be analyzed via straightforward linear methods: we can exactly characterize the linear contribution of each input to the model predictions; we can use a change-of-basis to disentangle input, output, and computational hidden unit subspaces; we can fully "},"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":"1611.09434","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-11-28T23:48:41Z","cross_cats_sorted":["cs.CL","cs.LG","cs.NE"],"title_canon_sha256":"19f6e9fd4fe90b85cd270710d45d0616d715a16704624ef9ea3bf816cd464cba","abstract_canon_sha256":"3b2f0cd3ace2472356445792b5b11f0800bece28815d73c809d143aaae1c36b8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:42:30.326297Z","signature_b64":"ec31Cm1dbzN+YtOsnTNVXwG5qO7Qb11yUwGHxz9ssoL6ZBUE6t5zNHImcsNVPbdICcgozwTD8XF5eaSZODI4DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c9b268ea141140cdcacd36225f63f0d185136462041ebd231b3c2ce18c067542","last_reissued_at":"2026-05-18T00:42:30.325741Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:42:30.325741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Input Switched Affine Networks: An RNN Architecture Designed for Interpretability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG","cs.NE"],"primary_cat":"cs.AI","authors_text":"David Sussillo, Jakob N. Foerster, Jan Chorowski, Jascha Sohl-Dickstein, Justin Gilmer","submitted_at":"2016-11-28T23:48:41Z","abstract_excerpt":"There exist many problem domains where the interpretability of neural network models is essential for deployment. Here we introduce a recurrent architecture composed of input-switched affine transformations - in other words an RNN without any explicit nonlinearities, but with input-dependent recurrent weights. This simple form allows the RNN to be analyzed via straightforward linear methods: we can exactly characterize the linear contribution of each input to the model predictions; we can use a change-of-basis to disentangle input, output, and computational hidden unit subspaces; we can fully "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1611.09434","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":""},"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":"1611.09434","created_at":"2026-05-18T00:42:30.325837+00:00"},{"alias_kind":"arxiv_version","alias_value":"1611.09434v2","created_at":"2026-05-18T00:42:30.325837+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1611.09434","created_at":"2026-05-18T00:42:30.325837+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZGZGR2QUCFAM","created_at":"2026-05-18T12:30:53.716459+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZGZGR2QUCFAM3SWN","created_at":"2026-05-18T12:30:53.716459+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZGZGR2QU","created_at":"2026-05-18T12:30:53.716459+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G","json":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G.json","graph_json":"https://pith.science/api/pith-number/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/graph.json","events_json":"https://pith.science/api/pith-number/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/events.json","paper":"https://pith.science/paper/ZGZGR2QU"},"agent_actions":{"view_html":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G","download_json":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G.json","view_paper":"https://pith.science/paper/ZGZGR2QU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1611.09434&json=true","fetch_graph":"https://pith.science/api/pith-number/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/graph.json","fetch_events":"https://pith.science/api/pith-number/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/action/storage_attestation","attest_author":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/action/author_attestation","sign_citation":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/action/citation_signature","submit_replication":"https://pith.science/pith/ZGZGR2QUCFAM3SWNGYRF6Y7Q2G/action/replication_record"}},"created_at":"2026-05-18T00:42:30.325837+00:00","updated_at":"2026-05-18T00:42:30.325837+00:00"}