{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:RRO25HFKDWRIV6VTJIJM4DJSQF","short_pith_number":"pith:RRO25HFK","schema_version":"1.0","canonical_sha256":"8c5dae9caa1da28afab34a12ce0d32816dafe4ae24ac498f0ba0204f6fd46239","source":{"kind":"arxiv","id":"2104.07085","version":4},"attestation_state":"computed","paper":{"title":"Fast Walsh-Hadamard Transform and Smooth-Thresholding Based Binary Layers in Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["eess.IV"],"primary_cat":"cs.CV","authors_text":"Ahmet Enis Cetin, Diaa Dabawi, Hongyi Pan","submitted_at":"2021-04-14T19:23:36Z","abstract_excerpt":"In this paper, we propose a novel layer based on fast Walsh-Hadamard transform (WHT) and smooth-thresholding to replace $1\\times 1$ convolution layers in deep neural networks. In the WHT domain, we denoise the transform domain coefficients using the new smooth-thresholding non-linearity, a smoothed version of the well-known soft-thresholding operator. We also introduce a family of multiplication-free operators from the basic 2$\\times$2 Hadamard transform to implement $3\\times 3$ depthwise separable convolution layers. Using these two types of layers, we replace the bottleneck layers in MobileN"},"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":"2104.07085","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-04-14T19:23:36Z","cross_cats_sorted":["eess.IV"],"title_canon_sha256":"10ef30d20a7ae6dc0819cd4c74ab547e9ef38c89f7d1c9c57d45e0e88258ceb0","abstract_canon_sha256":"09f4617d595cf3c2f62739636f04918a00f61d6647551e47e89429d18f2d0140"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:27:30.518571Z","signature_b64":"4tiF0PDOLIlHqVfQYU+YoxxEXAtXqNF2mEfujH4qyXSuU4z/biS4mrYqF2593HBN/5MX4BMqhV+UIrNt1i/ICw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8c5dae9caa1da28afab34a12ce0d32816dafe4ae24ac498f0ba0204f6fd46239","last_reissued_at":"2026-07-05T03:27:30.518055Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:27:30.518055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fast Walsh-Hadamard Transform and Smooth-Thresholding Based Binary Layers in Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["eess.IV"],"primary_cat":"cs.CV","authors_text":"Ahmet Enis Cetin, Diaa Dabawi, Hongyi Pan","submitted_at":"2021-04-14T19:23:36Z","abstract_excerpt":"In this paper, we propose a novel layer based on fast Walsh-Hadamard transform (WHT) and smooth-thresholding to replace $1\\times 1$ convolution layers in deep neural networks. In the WHT domain, we denoise the transform domain coefficients using the new smooth-thresholding non-linearity, a smoothed version of the well-known soft-thresholding operator. We also introduce a family of multiplication-free operators from the basic 2$\\times$2 Hadamard transform to implement $3\\times 3$ depthwise separable convolution layers. Using these two types of layers, we replace the bottleneck layers in MobileN"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.07085","kind":"arxiv","version":4},"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/2104.07085/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":"2104.07085","created_at":"2026-07-05T03:27:30.518117+00:00"},{"alias_kind":"arxiv_version","alias_value":"2104.07085v4","created_at":"2026-07-05T03:27:30.518117+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.07085","created_at":"2026-07-05T03:27:30.518117+00:00"},{"alias_kind":"pith_short_12","alias_value":"RRO25HFKDWRI","created_at":"2026-07-05T03:27:30.518117+00:00"},{"alias_kind":"pith_short_16","alias_value":"RRO25HFKDWRIV6VT","created_at":"2026-07-05T03:27:30.518117+00:00"},{"alias_kind":"pith_short_8","alias_value":"RRO25HFK","created_at":"2026-07-05T03:27:30.518117+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.09505","citing_title":"Combating the Memory Walls: Optimization Pathways for Long-Context Agentic LLM Inference","ref_index":54,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF","json":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF.json","graph_json":"https://pith.science/api/pith-number/RRO25HFKDWRIV6VTJIJM4DJSQF/graph.json","events_json":"https://pith.science/api/pith-number/RRO25HFKDWRIV6VTJIJM4DJSQF/events.json","paper":"https://pith.science/paper/RRO25HFK"},"agent_actions":{"view_html":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF","download_json":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF.json","view_paper":"https://pith.science/paper/RRO25HFK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2104.07085&json=true","fetch_graph":"https://pith.science/api/pith-number/RRO25HFKDWRIV6VTJIJM4DJSQF/graph.json","fetch_events":"https://pith.science/api/pith-number/RRO25HFKDWRIV6VTJIJM4DJSQF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF/action/storage_attestation","attest_author":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF/action/author_attestation","sign_citation":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF/action/citation_signature","submit_replication":"https://pith.science/pith/RRO25HFKDWRIV6VTJIJM4DJSQF/action/replication_record"}},"created_at":"2026-07-05T03:27:30.518117+00:00","updated_at":"2026-07-05T03:27:30.518117+00:00"}