{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:JHT44ZI7R3V4I6NBCGQVI2G424","short_pith_number":"pith:JHT44ZI7","schema_version":"1.0","canonical_sha256":"49e7ce651f8eebc479a111a15468dcd72cec7d8bed6297efb67dd603c472af2c","source":{"kind":"arxiv","id":"1602.02830","version":3},"attestation_state":"computed","paper":{"title":"Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel Soudry, Itay Hubara, Matthieu Courbariaux, Ran El-Yaniv, Yoshua Bengio","submitted_at":"2016-02-09T01:01:59Z","abstract_excerpt":"We introduce a method to train Binarized Neural Networks (BNNs) - neural networks with binary weights and activations at run-time. At training-time the binary weights and activations are used for computing the parameters gradients. During the forward pass, BNNs drastically reduce memory size and accesses, and replace most arithmetic operations with bit-wise operations, which is expected to substantially improve power-efficiency. To validate the effectiveness of BNNs we conduct two sets of experiments on the Torch7 and Theano frameworks. On both, BNNs achieved nearly state-of-the-art results ov"},"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":"1602.02830","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-02-09T01:01:59Z","cross_cats_sorted":[],"title_canon_sha256":"e1845ad7fcfe46ee574bc7d4c6bbd237b53a3790df044c3a176eb61ef16d9c79","abstract_canon_sha256":"f46186f7908a68069d165b667af957f20afc02fa3db170dd19c4978c6b1ad831"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:18:56.975670Z","signature_b64":"54ksttwf02mnsnEMpLknyasWMZ1tplaQFR/uEz+fT8IzzhMyhmsNrC/Qon9nQ2doSPsLwDG9DKkJpUMB9HMMDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"49e7ce651f8eebc479a111a15468dcd72cec7d8bed6297efb67dd603c472af2c","last_reissued_at":"2026-05-18T01:18:56.975056Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:18:56.975056Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Binarized Neural Networks: Training Deep Neural Networks with Weights and Activations Constrained to +1 or -1","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel Soudry, Itay Hubara, Matthieu Courbariaux, Ran El-Yaniv, Yoshua Bengio","submitted_at":"2016-02-09T01:01:59Z","abstract_excerpt":"We introduce a method to train Binarized Neural Networks (BNNs) - neural networks with binary weights and activations at run-time. At training-time the binary weights and activations are used for computing the parameters gradients. During the forward pass, BNNs drastically reduce memory size and accesses, and replace most arithmetic operations with bit-wise operations, which is expected to substantially improve power-efficiency. To validate the effectiveness of BNNs we conduct two sets of experiments on the Torch7 and Theano frameworks. On both, BNNs achieved nearly state-of-the-art results ov"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1602.02830","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":""},"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":"1602.02830","created_at":"2026-05-18T01:18:56.975132+00:00"},{"alias_kind":"arxiv_version","alias_value":"1602.02830v3","created_at":"2026-05-18T01:18:56.975132+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1602.02830","created_at":"2026-05-18T01:18:56.975132+00:00"},{"alias_kind":"pith_short_12","alias_value":"JHT44ZI7R3V4","created_at":"2026-05-18T12:30:25.849896+00:00"},{"alias_kind":"pith_short_16","alias_value":"JHT44ZI7R3V4I6NB","created_at":"2026-05-18T12:30:25.849896+00:00"},{"alias_kind":"pith_short_8","alias_value":"JHT44ZI7","created_at":"2026-05-18T12:30:25.849896+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":18,"internal_anchor_count":11,"sample":[{"citing_arxiv_id":"1906.09889","citing_title":"Improving Branch Prediction By Modeling Global History with Convolutional Neural Networks","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"1906.09395","citing_title":"Adaptive Precision CNN Accelerator Using Radix-X Parallel Connected Memristor Crossbars","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"1906.12172","citing_title":"New pointwise convolution in Deep Neural Networks through Extremely Fast and Non Parametric Transforms","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"1907.09077","citing_title":"A Stochastic-Computing based Deep Learning Framework using Adiabatic Quantum-Flux-Parametron SuperconductingTechnology","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"1907.10159","citing_title":"Efficient Detection and Quantification of Timing Leaks with Neural Networks","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"1907.10804","citing_title":"Co-Evolutionary Compression for Unpaired Image Translation","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2212.08989","citing_title":"Deep learning applied to computational mechanics: A comprehensive review, state of the art, and the classics","ref_index":297,"is_internal_anchor":true},{"citing_arxiv_id":"2508.06974","citing_title":"Rethinking 1-bit Optimization Leveraging Pre-trained Large Language Models","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.21171","citing_title":"FTerViT: Fully Ternary Vision Transformer","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10989","citing_title":"SURGE: Surrogate Gradient Adaptation in Binary Neural Networks","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2507.16079","citing_title":"A Lower Bound for the Number of Linear Regions of Ternary ReLU Regression Neural Networks","ref_index":1,"is_internal_anchor":true},{"citing_arxiv_id":"2208.07339","citing_title":"LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale","ref_index":125,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10989","citing_title":"SURGE: Surrogate Gradient Adaptation in Binary Neural Networks","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11558","citing_title":"A Composite Activation Function for Learning Stable Binary Representations","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03396","citing_title":"Design and Implementation of BNN-Based Object Detection on FPGA","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09604","citing_title":"DAP: Doppler-aware Point Network for Heterogeneous mmWave Action Recognition","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03396","citing_title":"Design and Implementation of BNN-Based Object Detection on FPGA","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19167","citing_title":"LBLLM: Lightweight Binarization of Large Language Models via Three-Stage Distillation","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424","json":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424.json","graph_json":"https://pith.science/api/pith-number/JHT44ZI7R3V4I6NBCGQVI2G424/graph.json","events_json":"https://pith.science/api/pith-number/JHT44ZI7R3V4I6NBCGQVI2G424/events.json","paper":"https://pith.science/paper/JHT44ZI7"},"agent_actions":{"view_html":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424","download_json":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424.json","view_paper":"https://pith.science/paper/JHT44ZI7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1602.02830&json=true","fetch_graph":"https://pith.science/api/pith-number/JHT44ZI7R3V4I6NBCGQVI2G424/graph.json","fetch_events":"https://pith.science/api/pith-number/JHT44ZI7R3V4I6NBCGQVI2G424/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424/action/storage_attestation","attest_author":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424/action/author_attestation","sign_citation":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424/action/citation_signature","submit_replication":"https://pith.science/pith/JHT44ZI7R3V4I6NBCGQVI2G424/action/replication_record"}},"created_at":"2026-05-18T01:18:56.975132+00:00","updated_at":"2026-05-18T01:18:56.975132+00:00"}