{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:ELKHASEXRNDRZVLY7XIC5R3MPP","short_pith_number":"pith:ELKHASEX","schema_version":"1.0","canonical_sha256":"22d47048978b471cd578fdd02ec76c7bf1af76c76cc4d033a12de2888d30f650","source":{"kind":"arxiv","id":"1812.08718","version":2},"attestation_state":"computed","paper":{"title":"RNNs Implicitly Implement Tensor Product Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ewan Dunbar, Paul Smolensky, R. Thomas McCoy, Tal Linzen","submitted_at":"2018-12-20T17:44:05Z","abstract_excerpt":"Recurrent neural networks (RNNs) can learn continuous vector representations of symbolic structures such as sequences and sentences; these representations often exhibit linear regularities (analogies). Such regularities motivate our hypothesis that RNNs that show such regularities implicitly compile symbolic structures into tensor product representations (TPRs; Smolensky, 1990), which additively combine tensor products of vectors representing roles (e.g., sequence positions) and vectors representing fillers (e.g., particular words). To test this hypothesis, we introduce Tensor Product Decompos"},"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":"1812.08718","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-12-20T17:44:05Z","cross_cats_sorted":[],"title_canon_sha256":"f7f54ae71147b3f56b43df86a575f569b1070da15c544974b9eb731f1d97fde2","abstract_canon_sha256":"2c5c944abf4271d7268e929cee7f927f13815b568b6b8e9a96dce9ddf5881251"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:52:03.586413Z","signature_b64":"JY8O2rrYFvzMoyKuGu7wuJo9kq+gVFZAQdDY0+JMT8l/eSQMqmfXylMaIVtHq37MC6SaTXkb0IZwt9QXsJnBBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"22d47048978b471cd578fdd02ec76c7bf1af76c76cc4d033a12de2888d30f650","last_reissued_at":"2026-05-17T23:52:03.585502Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:52:03.585502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RNNs Implicitly Implement Tensor Product Representations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ewan Dunbar, Paul Smolensky, R. Thomas McCoy, Tal Linzen","submitted_at":"2018-12-20T17:44:05Z","abstract_excerpt":"Recurrent neural networks (RNNs) can learn continuous vector representations of symbolic structures such as sequences and sentences; these representations often exhibit linear regularities (analogies). Such regularities motivate our hypothesis that RNNs that show such regularities implicitly compile symbolic structures into tensor product representations (TPRs; Smolensky, 1990), which additively combine tensor products of vectors representing roles (e.g., sequence positions) and vectors representing fillers (e.g., particular words). To test this hypothesis, we introduce Tensor Product Decompos"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.08718","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":"1812.08718","created_at":"2026-05-17T23:52:03.585608+00:00"},{"alias_kind":"arxiv_version","alias_value":"1812.08718v2","created_at":"2026-05-17T23:52:03.585608+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.08718","created_at":"2026-05-17T23:52:03.585608+00:00"},{"alias_kind":"pith_short_12","alias_value":"ELKHASEXRNDR","created_at":"2026-05-18T12:32:22.470017+00:00"},{"alias_kind":"pith_short_16","alias_value":"ELKHASEXRNDRZVLY","created_at":"2026-05-18T12:32:22.470017+00:00"},{"alias_kind":"pith_short_8","alias_value":"ELKHASEX","created_at":"2026-05-18T12:32:22.470017+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.23108","citing_title":"Artificial Phantasia: Emergent Mental Imagery in Large Language Models","ref_index":49,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP","json":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP.json","graph_json":"https://pith.science/api/pith-number/ELKHASEXRNDRZVLY7XIC5R3MPP/graph.json","events_json":"https://pith.science/api/pith-number/ELKHASEXRNDRZVLY7XIC5R3MPP/events.json","paper":"https://pith.science/paper/ELKHASEX"},"agent_actions":{"view_html":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP","download_json":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP.json","view_paper":"https://pith.science/paper/ELKHASEX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1812.08718&json=true","fetch_graph":"https://pith.science/api/pith-number/ELKHASEXRNDRZVLY7XIC5R3MPP/graph.json","fetch_events":"https://pith.science/api/pith-number/ELKHASEXRNDRZVLY7XIC5R3MPP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP/action/storage_attestation","attest_author":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP/action/author_attestation","sign_citation":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP/action/citation_signature","submit_replication":"https://pith.science/pith/ELKHASEXRNDRZVLY7XIC5R3MPP/action/replication_record"}},"created_at":"2026-05-17T23:52:03.585608+00:00","updated_at":"2026-05-17T23:52:03.585608+00:00"}