{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:FDDLIKUK3YF54SR6DEIIG72ACS","short_pith_number":"pith:FDDLIKUK","schema_version":"1.0","canonical_sha256":"28c6b42a8ade0bde4a3e1910837f401482116bee14c1489374380faf9ef38592","source":{"kind":"arxiv","id":"2309.01826","version":2},"attestation_state":"computed","paper":{"title":"One Wide Feedforward is All You Need","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ant\\'onio V. Lopes, Hendra Setiawan, Telmo Pessoa Pires, Yannick Assogba","submitted_at":"2023-09-04T21:30:21Z","abstract_excerpt":"The Transformer architecture has two main non-embedding components: Attention and the Feed Forward Network (FFN). Attention captures interdependencies between words regardless of their position, while the FFN non-linearly transforms each input token independently. In this work we explore the role of the FFN, and find that despite taking up a significant fraction of the model's parameters, it is highly redundant. Concretely, we are able to substantially reduce the number of parameters with only a modest drop in accuracy by removing the FFN on the decoder layers and sharing a single FFN across t"},"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":"2309.01826","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-04T21:30:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"001182d7b223ebecaedf3cd925e897bd953a4781c04272aa4d5b5e74ab446590","abstract_canon_sha256":"2994f983dfca421c35155689b31455ef96ebe79876724e38847a3d2d20f2f9e9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:03:30.641426Z","signature_b64":"zeX3GKUj5QhX/kPq0hW/Tewrvg3eJHnZqNzZYs2u3v+FImUnlCauzUYBEclBhHD0ReEbAmk5bSKtgIJK3OsRCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28c6b42a8ade0bde4a3e1910837f401482116bee14c1489374380faf9ef38592","last_reissued_at":"2026-07-05T07:03:30.640862Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:03:30.640862Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"One Wide Feedforward is All You Need","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ant\\'onio V. Lopes, Hendra Setiawan, Telmo Pessoa Pires, Yannick Assogba","submitted_at":"2023-09-04T21:30:21Z","abstract_excerpt":"The Transformer architecture has two main non-embedding components: Attention and the Feed Forward Network (FFN). Attention captures interdependencies between words regardless of their position, while the FFN non-linearly transforms each input token independently. In this work we explore the role of the FFN, and find that despite taking up a significant fraction of the model's parameters, it is highly redundant. Concretely, we are able to substantially reduce the number of parameters with only a modest drop in accuracy by removing the FFN on the decoder layers and sharing a single FFN across t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.01826","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/2309.01826/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":"2309.01826","created_at":"2026-07-05T07:03:30.640925+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.01826v2","created_at":"2026-07-05T07:03:30.640925+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.01826","created_at":"2026-07-05T07:03:30.640925+00:00"},{"alias_kind":"pith_short_12","alias_value":"FDDLIKUK3YF5","created_at":"2026-07-05T07:03:30.640925+00:00"},{"alias_kind":"pith_short_16","alias_value":"FDDLIKUK3YF54SR6","created_at":"2026-07-05T07:03:30.640925+00:00"},{"alias_kind":"pith_short_8","alias_value":"FDDLIKUK","created_at":"2026-07-05T07:03:30.640925+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2501.01802","citing_title":"BERT4MIMO: A Foundation Model using BERT Architecture for Massive MIMO Channel State Information Prediction","ref_index":32,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS","json":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS.json","graph_json":"https://pith.science/api/pith-number/FDDLIKUK3YF54SR6DEIIG72ACS/graph.json","events_json":"https://pith.science/api/pith-number/FDDLIKUK3YF54SR6DEIIG72ACS/events.json","paper":"https://pith.science/paper/FDDLIKUK"},"agent_actions":{"view_html":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS","download_json":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS.json","view_paper":"https://pith.science/paper/FDDLIKUK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.01826&json=true","fetch_graph":"https://pith.science/api/pith-number/FDDLIKUK3YF54SR6DEIIG72ACS/graph.json","fetch_events":"https://pith.science/api/pith-number/FDDLIKUK3YF54SR6DEIIG72ACS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS/action/storage_attestation","attest_author":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS/action/author_attestation","sign_citation":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS/action/citation_signature","submit_replication":"https://pith.science/pith/FDDLIKUK3YF54SR6DEIIG72ACS/action/replication_record"}},"created_at":"2026-07-05T07:03:30.640925+00:00","updated_at":"2026-07-05T07:03:30.640925+00:00"}