{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:3DG6VYYAAS2WJ2G64SYIEIGXII","short_pith_number":"pith:3DG6VYYA","schema_version":"1.0","canonical_sha256":"d8cdeae30004b564e8dee4b08220d7421d0c2d30a5d73544abee0b031582c96b","source":{"kind":"arxiv","id":"2305.17190","version":2},"attestation_state":"computed","paper":{"title":"Multiplication-Free Transformer Training via Piecewise Affine Operations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Atli Kosson, Martin Jaggi","submitted_at":"2023-05-26T18:28:28Z","abstract_excerpt":"Multiplications are responsible for most of the computational cost involved in neural network training and inference. Recent research has thus looked for ways to reduce the cost associated with them. Inspired by Mogami (2020), we replace multiplication with a cheap piecewise affine approximation that is achieved by adding the bit representation of the floating point numbers together as integers. We show that transformers can be trained with the resulting modified matrix multiplications on both vision and language tasks with little to no performance impact, and without changes to the training h"},"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":"2305.17190","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-05-26T18:28:28Z","cross_cats_sorted":[],"title_canon_sha256":"34013005c1c7110a1b2dee3cd78e41f3c7c5f6c1a5a21087ea62dc9e077d461c","abstract_canon_sha256":"be580b4a7995db17ec427a1c17a0163da1f95e726df3dfa5cc0c9d465d113171"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:04:47.734784Z","signature_b64":"K3XCHwPi4EuG13kshpRuYQCiI0y066/ZZWQmCVTYCvoqpWydLgqJWIRIuHBTI93E8nroi6K/o/VxbgT0hJfbCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d8cdeae30004b564e8dee4b08220d7421d0c2d30a5d73544abee0b031582c96b","last_reissued_at":"2026-07-05T07:04:47.734389Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:04:47.734389Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multiplication-Free Transformer Training via Piecewise Affine Operations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Atli Kosson, Martin Jaggi","submitted_at":"2023-05-26T18:28:28Z","abstract_excerpt":"Multiplications are responsible for most of the computational cost involved in neural network training and inference. Recent research has thus looked for ways to reduce the cost associated with them. Inspired by Mogami (2020), we replace multiplication with a cheap piecewise affine approximation that is achieved by adding the bit representation of the floating point numbers together as integers. We show that transformers can be trained with the resulting modified matrix multiplications on both vision and language tasks with little to no performance impact, and without changes to the training h"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.17190","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/2305.17190/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":"2305.17190","created_at":"2026-07-05T07:04:47.734446+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.17190v2","created_at":"2026-07-05T07:04:47.734446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.17190","created_at":"2026-07-05T07:04:47.734446+00:00"},{"alias_kind":"pith_short_12","alias_value":"3DG6VYYAAS2W","created_at":"2026-07-05T07:04:47.734446+00:00"},{"alias_kind":"pith_short_16","alias_value":"3DG6VYYAAS2WJ2G6","created_at":"2026-07-05T07:04:47.734446+00:00"},{"alias_kind":"pith_short_8","alias_value":"3DG6VYYA","created_at":"2026-07-05T07:04:47.734446+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2503.18002","citing_title":"Neuromorphic Principles for Efficient Large Language Models on Intel Loihi 2","ref_index":14,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII","json":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII.json","graph_json":"https://pith.science/api/pith-number/3DG6VYYAAS2WJ2G64SYIEIGXII/graph.json","events_json":"https://pith.science/api/pith-number/3DG6VYYAAS2WJ2G64SYIEIGXII/events.json","paper":"https://pith.science/paper/3DG6VYYA"},"agent_actions":{"view_html":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII","download_json":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII.json","view_paper":"https://pith.science/paper/3DG6VYYA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.17190&json=true","fetch_graph":"https://pith.science/api/pith-number/3DG6VYYAAS2WJ2G64SYIEIGXII/graph.json","fetch_events":"https://pith.science/api/pith-number/3DG6VYYAAS2WJ2G64SYIEIGXII/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII/action/storage_attestation","attest_author":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII/action/author_attestation","sign_citation":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII/action/citation_signature","submit_replication":"https://pith.science/pith/3DG6VYYAAS2WJ2G64SYIEIGXII/action/replication_record"}},"created_at":"2026-07-05T07:04:47.734446+00:00","updated_at":"2026-07-05T07:04:47.734446+00:00"}