{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:AWK5SVBKSDSBPA2MHILMVBTYOL","short_pith_number":"pith:AWK5SVBK","canonical_record":{"source":{"id":"1808.02513","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-08-07T18:42:09Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"e9e908731c788063118a90eea8547efbdae33a0ba0bcae124334c864eed206ba","abstract_canon_sha256":"c4f48c32da08ad18a86306f35484895cd3ca4de53739bd4fd6cefffb710a7351"},"schema_version":"1.0"},"canonical_sha256":"0595d9542a90e417834c3a16ca867872c5ecaa0d46515adc37a9680309242d0c","source":{"kind":"arxiv","id":"1808.02513","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.02513","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"arxiv_version","alias_value":"1808.02513v1","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.02513","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"pith_short_12","alias_value":"AWK5SVBKSDSB","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_16","alias_value":"AWK5SVBKSDSBPA2M","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_8","alias_value":"AWK5SVBK","created_at":"2026-05-18T12:32:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:AWK5SVBKSDSBPA2MHILMVBTYOL","target":"record","payload":{"canonical_record":{"source":{"id":"1808.02513","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-08-07T18:42:09Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"e9e908731c788063118a90eea8547efbdae33a0ba0bcae124334c864eed206ba","abstract_canon_sha256":"c4f48c32da08ad18a86306f35484895cd3ca4de53739bd4fd6cefffb710a7351"},"schema_version":"1.0"},"canonical_sha256":"0595d9542a90e417834c3a16ca867872c5ecaa0d46515adc37a9680309242d0c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:08:35.203413Z","signature_b64":"aTWbWrL6tto/5nkT0nY5UsWy6OTsutcZBxrDQIwwtvTOyI2Eqg0S03dCsRHkU1h9n0TMDYcpsP9aYlnvEbsCCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0595d9542a90e417834c3a16ca867872c5ecaa0d46515adc37a9680309242d0c","last_reissued_at":"2026-05-18T00:08:35.202754Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:08:35.202754Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1808.02513","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:08:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SKhk+NiUq4RIjNSglSFcSWIUghZAiR7DF7KD3bmgUuRisE0XsLKHJSywMmOq9zfV2LdS0q38WOAAuXapDbbDAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-07T18:26:14.703479Z"},"content_sha256":"e662f23b129e963316517d5138ae37c249c39d1dbfa97e7b2a5bf3acccf60114","schema_version":"1.0","event_id":"sha256:e662f23b129e963316517d5138ae37c249c39d1dbfa97e7b2a5bf3acccf60114"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:AWK5SVBKSDSBPA2MHILMVBTYOL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Rethinking Numerical Representations for Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Babak Zamirai, Jason Mars, Jia Deng, Lingjia Tang, Marios Papaefthymiou, Mehrzad Samadi, Michael Laurenzano, Parker Hill, Scott Mahlke, Shengshuo Lu, Thomas Wenisch, Yu-Wei Chao","submitted_at":"2018-08-07T18:42:09Z","abstract_excerpt":"With ever-increasing computational demand for deep learning, it is critical to investigate the implications of the numeric representation and precision of DNN model weights and activations on computational efficiency. In this work, we explore unconventional narrow-precision floating-point representations as it relates to inference accuracy and efficiency to steer the improved design of future DNN platforms. We show that inference using these custom numeric representations on production-grade DNNs, including GoogLeNet and VGG, achieves an average speedup of 7.6x with less than 1% degradation in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.02513","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":""},"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":"2026-05-18T00:08:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EfgCQi5l3JNmx/29fw9uCETsWA0kifj+bTPLShcvMIDSd+5a3+RafpHCf0LVnUDN8STf0qJz6HRmJ9PFzPL7CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-07T18:26:14.704140Z"},"content_sha256":"3c9c4978a560d4a124c1fdd24b6fd847ffcacdd7f94f4137cd48997aec5b0048","schema_version":"1.0","event_id":"sha256:3c9c4978a560d4a124c1fdd24b6fd847ffcacdd7f94f4137cd48997aec5b0048"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/bundle.json","state_url":"https://pith.science/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/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":"2026-06-07T18:26:14Z","links":{"resolver":"https://pith.science/pith/AWK5SVBKSDSBPA2MHILMVBTYOL","bundle":"https://pith.science/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/bundle.json","state":"https://pith.science/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AWK5SVBKSDSBPA2MHILMVBTYOL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:AWK5SVBKSDSBPA2MHILMVBTYOL","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":"c4f48c32da08ad18a86306f35484895cd3ca4de53739bd4fd6cefffb710a7351","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-08-07T18:42:09Z","title_canon_sha256":"e9e908731c788063118a90eea8547efbdae33a0ba0bcae124334c864eed206ba"},"schema_version":"1.0","source":{"id":"1808.02513","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1808.02513","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"arxiv_version","alias_value":"1808.02513v1","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1808.02513","created_at":"2026-05-18T00:08:35Z"},{"alias_kind":"pith_short_12","alias_value":"AWK5SVBKSDSB","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_16","alias_value":"AWK5SVBKSDSBPA2M","created_at":"2026-05-18T12:32:13Z"},{"alias_kind":"pith_short_8","alias_value":"AWK5SVBK","created_at":"2026-05-18T12:32:13Z"}],"graph_snapshots":[{"event_id":"sha256:3c9c4978a560d4a124c1fdd24b6fd847ffcacdd7f94f4137cd48997aec5b0048","target":"graph","created_at":"2026-05-18T00:08:35Z","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":"With ever-increasing computational demand for deep learning, it is critical to investigate the implications of the numeric representation and precision of DNN model weights and activations on computational efficiency. In this work, we explore unconventional narrow-precision floating-point representations as it relates to inference accuracy and efficiency to steer the improved design of future DNN platforms. We show that inference using these custom numeric representations on production-grade DNNs, including GoogLeNet and VGG, achieves an average speedup of 7.6x with less than 1% degradation in","authors_text":"Babak Zamirai, Jason Mars, Jia Deng, Lingjia Tang, Marios Papaefthymiou, Mehrzad Samadi, Michael Laurenzano, Parker Hill, Scott Mahlke, Shengshuo Lu, Thomas Wenisch, Yu-Wei Chao","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-08-07T18:42:09Z","title":"Rethinking Numerical Representations for Deep Neural Networks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1808.02513","kind":"arxiv","version":1},"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:e662f23b129e963316517d5138ae37c249c39d1dbfa97e7b2a5bf3acccf60114","target":"record","created_at":"2026-05-18T00:08:35Z","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":"c4f48c32da08ad18a86306f35484895cd3ca4de53739bd4fd6cefffb710a7351","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-08-07T18:42:09Z","title_canon_sha256":"e9e908731c788063118a90eea8547efbdae33a0ba0bcae124334c864eed206ba"},"schema_version":"1.0","source":{"id":"1808.02513","kind":"arxiv","version":1}},"canonical_sha256":"0595d9542a90e417834c3a16ca867872c5ecaa0d46515adc37a9680309242d0c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0595d9542a90e417834c3a16ca867872c5ecaa0d46515adc37a9680309242d0c","first_computed_at":"2026-05-18T00:08:35.202754Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:08:35.202754Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aTWbWrL6tto/5nkT0nY5UsWy6OTsutcZBxrDQIwwtvTOyI2Eqg0S03dCsRHkU1h9n0TMDYcpsP9aYlnvEbsCCw==","signature_status":"signed_v1","signed_at":"2026-05-18T00:08:35.203413Z","signed_message":"canonical_sha256_bytes"},"source_id":"1808.02513","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e662f23b129e963316517d5138ae37c249c39d1dbfa97e7b2a5bf3acccf60114","sha256:3c9c4978a560d4a124c1fdd24b6fd847ffcacdd7f94f4137cd48997aec5b0048"],"state_sha256":"c5d71793e58bc1bb40e5e790fef9a7c70befebba8c304ac6ce59c4d3ff8bfc78"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"BEY2l/RHIjCa2mWJw7XjxWh4RfwaEFKz8uq47KovHKWWW+Aof2whsFI0JXUp+do+YHVoO7gYVQTgmgF6Tu0WCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-07T18:26:14.708545Z","bundle_sha256":"901170ebb6e29fd11059efa1f591a7d7f4569431671a0fad666be1a01d235466"}}