{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2014:B26NHBQ4AU22TGS4A7E55X6F7Z","short_pith_number":"pith:B26NHBQ4","canonical_record":{"source":{"id":"1412.7024","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2014-12-22T15:22:45Z","cross_cats_sorted":["cs.CV","cs.NE"],"title_canon_sha256":"daf9b2ef3c266d97574571c2d4b2035f0e258b9f1c6149adbaf4a450140935f1","abstract_canon_sha256":"985abeae64dad80f835a0bc4098672dd34907c43dbe4bd535d7fab5ffdcb2d4a"},"schema_version":"1.0"},"canonical_sha256":"0ebcd3861c0535a99a5c07c9dedfc5fe60192e2fd4a84ee6569e3b7d49d4c637","source":{"kind":"arxiv","id":"1412.7024","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1412.7024","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"arxiv_version","alias_value":"1412.7024v5","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1412.7024","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"pith_short_12","alias_value":"B26NHBQ4AU22","created_at":"2026-05-18T12:28:19Z"},{"alias_kind":"pith_short_16","alias_value":"B26NHBQ4AU22TGS4","created_at":"2026-05-18T12:28:19Z"},{"alias_kind":"pith_short_8","alias_value":"B26NHBQ4","created_at":"2026-05-18T12:28:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2014:B26NHBQ4AU22TGS4A7E55X6F7Z","target":"record","payload":{"canonical_record":{"source":{"id":"1412.7024","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2014-12-22T15:22:45Z","cross_cats_sorted":["cs.CV","cs.NE"],"title_canon_sha256":"daf9b2ef3c266d97574571c2d4b2035f0e258b9f1c6149adbaf4a450140935f1","abstract_canon_sha256":"985abeae64dad80f835a0bc4098672dd34907c43dbe4bd535d7fab5ffdcb2d4a"},"schema_version":"1.0"},"canonical_sha256":"0ebcd3861c0535a99a5c07c9dedfc5fe60192e2fd4a84ee6569e3b7d49d4c637","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:32:18.826848Z","signature_b64":"VYE70i8A9GrEc66SMAKxdVfHQu5/gznck1Vnk2159u1q6qn9coorJC/o/YANsJPJEGuOsA9wNQK1e2t82gOFDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ebcd3861c0535a99a5c07c9dedfc5fe60192e2fd4a84ee6569e3b7d49d4c637","last_reissued_at":"2026-05-18T01:32:18.826194Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:32:18.826194Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1412.7024","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:32:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"exuqGI+PLHDpj7OQjAPAp8idhRKRUqICTah9TlJwqAxTJSlsnivhTdzt/w3a5DdtRB6cLg79jHQT/DS5HKF+AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T20:19:59.574215Z"},"content_sha256":"69e211632c356bd2b60dff5563fe1c6a33dac8b18240c131281a34d0385def8a","schema_version":"1.0","event_id":"sha256:69e211632c356bd2b60dff5563fe1c6a33dac8b18240c131281a34d0385def8a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2014:B26NHBQ4AU22TGS4A7E55X6F7Z","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Training deep neural networks with low precision multiplications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.NE"],"primary_cat":"cs.LG","authors_text":"Jean-Pierre David, Matthieu Courbariaux, Yoshua Bengio","submitted_at":"2014-12-22T15:22:45Z","abstract_excerpt":"Multipliers are the most space and power-hungry arithmetic operators of the digital implementation of deep neural networks. We train a set of state-of-the-art neural networks (Maxout networks) on three benchmark datasets: MNIST, CIFAR-10 and SVHN. They are trained with three distinct formats: floating point, fixed point and dynamic fixed point. For each of those datasets and for each of those formats, we assess the impact of the precision of the multiplications on the final error after training. We find that very low precision is sufficient not just for running trained networks but also for tr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1412.7024","kind":"arxiv","version":5},"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T01:32:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RaYnAUOAgNIf8ZU3X7639pqJ0Qe9b1hTmZpIFGlqTfhmBiUF2hiBY+NSasletOaYCOZxoB8Sxd0MvKcPWaOtAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T20:19:59.574933Z"},"content_sha256":"6a83c01d96457294abacee730fbc03c606e92b1dada21a7bf25f0f0d1e4c902b","schema_version":"1.0","event_id":"sha256:6a83c01d96457294abacee730fbc03c606e92b1dada21a7bf25f0f0d1e4c902b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/bundle.json","state_url":"https://pith.science/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-25T20:19:59Z","links":{"resolver":"https://pith.science/pith/B26NHBQ4AU22TGS4A7E55X6F7Z","bundle":"https://pith.science/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/bundle.json","state":"https://pith.science/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/state.json","well_known_bundle":"https://pith.science/.well-known/pith/B26NHBQ4AU22TGS4A7E55X6F7Z/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2014:B26NHBQ4AU22TGS4A7E55X6F7Z","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"985abeae64dad80f835a0bc4098672dd34907c43dbe4bd535d7fab5ffdcb2d4a","cross_cats_sorted":["cs.CV","cs.NE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2014-12-22T15:22:45Z","title_canon_sha256":"daf9b2ef3c266d97574571c2d4b2035f0e258b9f1c6149adbaf4a450140935f1"},"schema_version":"1.0","source":{"id":"1412.7024","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1412.7024","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"arxiv_version","alias_value":"1412.7024v5","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1412.7024","created_at":"2026-05-18T01:32:18Z"},{"alias_kind":"pith_short_12","alias_value":"B26NHBQ4AU22","created_at":"2026-05-18T12:28:19Z"},{"alias_kind":"pith_short_16","alias_value":"B26NHBQ4AU22TGS4","created_at":"2026-05-18T12:28:19Z"},{"alias_kind":"pith_short_8","alias_value":"B26NHBQ4","created_at":"2026-05-18T12:28:19Z"}],"graph_snapshots":[{"event_id":"sha256:6a83c01d96457294abacee730fbc03c606e92b1dada21a7bf25f0f0d1e4c902b","target":"graph","created_at":"2026-05-18T01:32:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Multipliers are the most space and power-hungry arithmetic operators of the digital implementation of deep neural networks. We train a set of state-of-the-art neural networks (Maxout networks) on three benchmark datasets: MNIST, CIFAR-10 and SVHN. They are trained with three distinct formats: floating point, fixed point and dynamic fixed point. For each of those datasets and for each of those formats, we assess the impact of the precision of the multiplications on the final error after training. We find that very low precision is sufficient not just for running trained networks but also for tr","authors_text":"Jean-Pierre David, Matthieu Courbariaux, Yoshua Bengio","cross_cats":["cs.CV","cs.NE"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2014-12-22T15:22:45Z","title":"Training deep neural networks with low precision multiplications"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1412.7024","kind":"arxiv","version":5},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:69e211632c356bd2b60dff5563fe1c6a33dac8b18240c131281a34d0385def8a","target":"record","created_at":"2026-05-18T01:32:18Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"985abeae64dad80f835a0bc4098672dd34907c43dbe4bd535d7fab5ffdcb2d4a","cross_cats_sorted":["cs.CV","cs.NE"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2014-12-22T15:22:45Z","title_canon_sha256":"daf9b2ef3c266d97574571c2d4b2035f0e258b9f1c6149adbaf4a450140935f1"},"schema_version":"1.0","source":{"id":"1412.7024","kind":"arxiv","version":5}},"canonical_sha256":"0ebcd3861c0535a99a5c07c9dedfc5fe60192e2fd4a84ee6569e3b7d49d4c637","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0ebcd3861c0535a99a5c07c9dedfc5fe60192e2fd4a84ee6569e3b7d49d4c637","first_computed_at":"2026-05-18T01:32:18.826194Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T01:32:18.826194Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VYE70i8A9GrEc66SMAKxdVfHQu5/gznck1Vnk2159u1q6qn9coorJC/o/YANsJPJEGuOsA9wNQK1e2t82gOFDg==","signature_status":"signed_v1","signed_at":"2026-05-18T01:32:18.826848Z","signed_message":"canonical_sha256_bytes"},"source_id":"1412.7024","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:69e211632c356bd2b60dff5563fe1c6a33dac8b18240c131281a34d0385def8a","sha256:6a83c01d96457294abacee730fbc03c606e92b1dada21a7bf25f0f0d1e4c902b"],"state_sha256":"a61d1a617906ba538d06465619b113656698e093552e1c2879f91581ce25cce2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oHeWwp8lZLvHlYLuG+5x5qIYXLfJSMnFxJf60daj8ICv5Bc4ifp9zEngX+hBAWtHYy0B8A55X36CuKDUNYGvCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T20:19:59.578330Z","bundle_sha256":"cd472301224a860a657716da73d59f9aec201d132fa6ca617e41070160d32a19"}}