{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:LWRZHIT6BXVSEMCJ4RSEZLQPGL","short_pith_number":"pith:LWRZHIT6","schema_version":"1.0","canonical_sha256":"5da393a27e0deb223049e4644cae0f32e89c0bfbe1ef1baf8b89522a882183a7","source":{"kind":"arxiv","id":"2303.03667","version":3},"attestation_state":"computed","paper":{"title":"Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chul-Ho Lee, Hao He, Jierun Chen, S.-H. Gary Chan, Shiu-hong Kao, Song Wen, Weipeng Zhuo","submitted_at":"2023-03-07T06:05:30Z","abstract_excerpt":"To design fast neural networks, many works have been focusing on reducing the number of floating-point operations (FLOPs). We observe that such reduction in FLOPs, however, does not necessarily lead to a similar level of reduction in latency. This mainly stems from inefficiently low floating-point operations per second (FLOPS). To achieve faster networks, we revisit popular operators and demonstrate that such low FLOPS is mainly due to frequent memory access of the operators, especially the depthwise convolution. We hence propose a novel partial convolution (PConv) that extracts spatial featur"},"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":"2303.03667","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-03-07T06:05:30Z","cross_cats_sorted":[],"title_canon_sha256":"835e9b87e3dc0257b9006492b03355db0bd7b0a35df6d41296ec97ac2409c000","abstract_canon_sha256":"3e9d5bf0512a64d789e3e90aa00080f845b3907fcadad5971a2327399916fbf0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:12:00.551904Z","signature_b64":"4t/j7aXcfWoXGFlPS6WaQEh+4jK5lXBE+3AzhYDNKu2pqodJ4hRayz7O+j6r0aP2hXQlnxdLxT/acgN/BSQ7Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5da393a27e0deb223049e4644cae0f32e89c0bfbe1ef1baf8b89522a882183a7","last_reissued_at":"2026-07-05T06:12:00.551440Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:12:00.551440Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chul-Ho Lee, Hao He, Jierun Chen, S.-H. Gary Chan, Shiu-hong Kao, Song Wen, Weipeng Zhuo","submitted_at":"2023-03-07T06:05:30Z","abstract_excerpt":"To design fast neural networks, many works have been focusing on reducing the number of floating-point operations (FLOPs). We observe that such reduction in FLOPs, however, does not necessarily lead to a similar level of reduction in latency. This mainly stems from inefficiently low floating-point operations per second (FLOPS). To achieve faster networks, we revisit popular operators and demonstrate that such low FLOPS is mainly due to frequent memory access of the operators, especially the depthwise convolution. We hence propose a novel partial convolution (PConv) that extracts spatial featur"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.03667","kind":"arxiv","version":3},"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/2303.03667/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":"2303.03667","created_at":"2026-07-05T06:12:00.551494+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.03667v3","created_at":"2026-07-05T06:12:00.551494+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.03667","created_at":"2026-07-05T06:12:00.551494+00:00"},{"alias_kind":"pith_short_12","alias_value":"LWRZHIT6BXVS","created_at":"2026-07-05T06:12:00.551494+00:00"},{"alias_kind":"pith_short_16","alias_value":"LWRZHIT6BXVSEMCJ","created_at":"2026-07-05T06:12:00.551494+00:00"},{"alias_kind":"pith_short_8","alias_value":"LWRZHIT6","created_at":"2026-07-05T06:12:00.551494+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.05490","citing_title":"A Weak-Signal-Aware Framework for Subsurface Defect Detection: Mechanisms for Enhancing Low-SCR Hyperbolic Signatures","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14884","citing_title":"FSDETR: Frequency-Spatial Feature Enhancement for Small Object Detection","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL","json":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL.json","graph_json":"https://pith.science/api/pith-number/LWRZHIT6BXVSEMCJ4RSEZLQPGL/graph.json","events_json":"https://pith.science/api/pith-number/LWRZHIT6BXVSEMCJ4RSEZLQPGL/events.json","paper":"https://pith.science/paper/LWRZHIT6"},"agent_actions":{"view_html":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL","download_json":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL.json","view_paper":"https://pith.science/paper/LWRZHIT6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.03667&json=true","fetch_graph":"https://pith.science/api/pith-number/LWRZHIT6BXVSEMCJ4RSEZLQPGL/graph.json","fetch_events":"https://pith.science/api/pith-number/LWRZHIT6BXVSEMCJ4RSEZLQPGL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL/action/storage_attestation","attest_author":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL/action/author_attestation","sign_citation":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL/action/citation_signature","submit_replication":"https://pith.science/pith/LWRZHIT6BXVSEMCJ4RSEZLQPGL/action/replication_record"}},"created_at":"2026-07-05T06:12:00.551494+00:00","updated_at":"2026-07-05T06:12:00.551494+00:00"}