{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:IOG4EQBFY5S2CA45D5UPF2WGCE","short_pith_number":"pith:IOG4EQBF","schema_version":"1.0","canonical_sha256":"438dc24025c765a1039d1f68f2eac6110d10a000477647990dc1b45934225b0f","source":{"kind":"arxiv","id":"2209.12982","version":1},"attestation_state":"computed","paper":{"title":"Going Further With Winograd Convolutions: Tap-Wise Quantization for Efficient Inference on 4x4 Tile","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.AR","authors_text":"Antonio Cipolletta, Beatrice Bussolino, Lukas Cavigelli, Renzo Andri, Zhe Wang","submitted_at":"2022-09-26T19:29:51Z","abstract_excerpt":"Most of today's computer vision pipelines are built around deep neural networks, where convolution operations require most of the generally high compute effort. The Winograd convolution algorithm computes convolutions with fewer MACs compared to the standard algorithm, reducing the operation count by a factor of 2.25x for 3x3 convolutions when using the version with 2x2-sized tiles $F_2$. Even though the gain is significant, the Winograd algorithm with larger tile sizes, i.e., $F_4$, offers even more potential in improving throughput and energy efficiency, as it reduces the required MACs by 4x"},"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":"2209.12982","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AR","submitted_at":"2022-09-26T19:29:51Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"b1c6d6cc83472f8b95f050885b97152d3ca2ad038fc04587e1cbaceb7169a30b","abstract_canon_sha256":"7a2e8105fabe8c4d15ef84a1ed6e4cf27ad0ea3241a4509311701fdaabe6f58e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:01:05.141937Z","signature_b64":"q0aofXl9YnLd8DfQBLgaNpk6F7HSF/Vw0uFch5n9mO55658NREoenrwaFktegvTei6Jrs3h8YvOywV1GheV0Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"438dc24025c765a1039d1f68f2eac6110d10a000477647990dc1b45934225b0f","last_reissued_at":"2026-07-05T05:01:05.141553Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:01:05.141553Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Going Further With Winograd Convolutions: Tap-Wise Quantization for Efficient Inference on 4x4 Tile","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.AR","authors_text":"Antonio Cipolletta, Beatrice Bussolino, Lukas Cavigelli, Renzo Andri, Zhe Wang","submitted_at":"2022-09-26T19:29:51Z","abstract_excerpt":"Most of today's computer vision pipelines are built around deep neural networks, where convolution operations require most of the generally high compute effort. The Winograd convolution algorithm computes convolutions with fewer MACs compared to the standard algorithm, reducing the operation count by a factor of 2.25x for 3x3 convolutions when using the version with 2x2-sized tiles $F_2$. Even though the gain is significant, the Winograd algorithm with larger tile sizes, i.e., $F_4$, offers even more potential in improving throughput and energy efficiency, as it reduces the required MACs by 4x"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.12982","kind":"arxiv","version":1},"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/2209.12982/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":"2209.12982","created_at":"2026-07-05T05:01:05.141609+00:00"},{"alias_kind":"arxiv_version","alias_value":"2209.12982v1","created_at":"2026-07-05T05:01:05.141609+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.12982","created_at":"2026-07-05T05:01:05.141609+00:00"},{"alias_kind":"pith_short_12","alias_value":"IOG4EQBFY5S2","created_at":"2026-07-05T05:01:05.141609+00:00"},{"alias_kind":"pith_short_16","alias_value":"IOG4EQBFY5S2CA45","created_at":"2026-07-05T05:01:05.141609+00:00"},{"alias_kind":"pith_short_8","alias_value":"IOG4EQBF","created_at":"2026-07-05T05:01:05.141609+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/IOG4EQBFY5S2CA45D5UPF2WGCE","json":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE.json","graph_json":"https://pith.science/api/pith-number/IOG4EQBFY5S2CA45D5UPF2WGCE/graph.json","events_json":"https://pith.science/api/pith-number/IOG4EQBFY5S2CA45D5UPF2WGCE/events.json","paper":"https://pith.science/paper/IOG4EQBF"},"agent_actions":{"view_html":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE","download_json":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE.json","view_paper":"https://pith.science/paper/IOG4EQBF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2209.12982&json=true","fetch_graph":"https://pith.science/api/pith-number/IOG4EQBFY5S2CA45D5UPF2WGCE/graph.json","fetch_events":"https://pith.science/api/pith-number/IOG4EQBFY5S2CA45D5UPF2WGCE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE/action/storage_attestation","attest_author":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE/action/author_attestation","sign_citation":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE/action/citation_signature","submit_replication":"https://pith.science/pith/IOG4EQBFY5S2CA45D5UPF2WGCE/action/replication_record"}},"created_at":"2026-07-05T05:01:05.141609+00:00","updated_at":"2026-07-05T05:01:05.141609+00:00"}