{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2017:WXZD7ITVZT2B2LRFUZMV4NR45L","short_pith_number":"pith:WXZD7ITV","canonical_record":{"source":{"id":"1711.02213","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-06T23:05:10Z","cross_cats_sorted":["cs.NA","stat.ML"],"title_canon_sha256":"13029903ba7673a21330464a22da8f64320910af64ac2597e2d603c4faf56ba5","abstract_canon_sha256":"344de0b9f05f4e6980704eb16e20ae6e355b3dcc10e2d0be0ea8d85b5f7dd9d8"},"schema_version":"1.0"},"canonical_sha256":"b5f23fa275ccf41d2e25a6595e363ceaeb2d0d1b17a2722127c3df249841ee46","source":{"kind":"arxiv","id":"1711.02213","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1711.02213","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"arxiv_version","alias_value":"1711.02213v2","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1711.02213","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"pith_short_12","alias_value":"WXZD7ITVZT2B","created_at":"2026-05-18T12:31:53Z"},{"alias_kind":"pith_short_16","alias_value":"WXZD7ITVZT2B2LRF","created_at":"2026-05-18T12:31:53Z"},{"alias_kind":"pith_short_8","alias_value":"WXZD7ITV","created_at":"2026-05-18T12:31:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2017:WXZD7ITVZT2B2LRFUZMV4NR45L","target":"record","payload":{"canonical_record":{"source":{"id":"1711.02213","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-06T23:05:10Z","cross_cats_sorted":["cs.NA","stat.ML"],"title_canon_sha256":"13029903ba7673a21330464a22da8f64320910af64ac2597e2d603c4faf56ba5","abstract_canon_sha256":"344de0b9f05f4e6980704eb16e20ae6e355b3dcc10e2d0be0ea8d85b5f7dd9d8"},"schema_version":"1.0"},"canonical_sha256":"b5f23fa275ccf41d2e25a6595e363ceaeb2d0d1b17a2722127c3df249841ee46","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:29:05.392829Z","signature_b64":"8PmTteSglD6TaGqDvSfA2GQSpEN/Z5D9FoT7SxLwMhy2sIe45kcU58lset5ucseMOFEX0xt0DPdM8QS2jKX7DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b5f23fa275ccf41d2e25a6595e363ceaeb2d0d1b17a2722127c3df249841ee46","last_reissued_at":"2026-05-18T00:29:05.392173Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:29:05.392173Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1711.02213","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:29:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hUfwNfwSMLjo8dr/8bKKQL2YWobLxBMctBMGeeAoASDQcfmcSj0aD6WIZK6CKzfAy7ozgkH6pFJYN9lMt/dNAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T10:57:42.240934Z"},"content_sha256":"296bc82164e720ec8c7a0d30de685015a866dffe2f2a346fc6d8d81f02818e5d","schema_version":"1.0","event_id":"sha256:296bc82164e720ec8c7a0d30de685015a866dffe2f2a346fc6d8d81f02818e5d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2017:WXZD7ITVZT2B2LRFUZMV4NR45L","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NA","stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Khosrowshahi, Arjun K. Bansal, Carey Kloss, Luke Hornof, Marcel Nassar, Naveen Rao, O\\u{g}uz H. Elibol, Ruby J. Pai, Scott Gray, Stewart Hall, Tristan J. Webb, Urs K\\\"oster, William H. Constable, Xin Wang","submitted_at":"2017-11-06T23:05:10Z","abstract_excerpt":"Deep neural networks are commonly developed and trained in 32-bit floating point format. Significant gains in performance and energy efficiency could be realized by training and inference in numerical formats optimized for deep learning. Despite advances in limited precision inference in recent years, training of neural networks in low bit-width remains a challenging problem. Here we present the Flexpoint data format, aiming at a complete replacement of 32-bit floating point format training and inference, designed to support modern deep network topologies without modifications. Flexpoint tenso"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1711.02213","kind":"arxiv","version":2},"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":"2026-05-18T00:29:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4X8PsORqj9/A6RMLpGbPs+YYx+lIBv5s1yGy4voFypLRZGyUJctkL3azNb9yNW3wGRVZzf6iBHqWvXHLDWFjAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T10:57:42.241498Z"},"content_sha256":"e11f917155faa6c77f584e181d8ae4db1851176cb194579d9ade99c503c7ddde","schema_version":"1.0","event_id":"sha256:e11f917155faa6c77f584e181d8ae4db1851176cb194579d9ade99c503c7ddde"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/bundle.json","state_url":"https://pith.science/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/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":"2026-05-30T10:57:42Z","links":{"resolver":"https://pith.science/pith/WXZD7ITVZT2B2LRFUZMV4NR45L","bundle":"https://pith.science/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/bundle.json","state":"https://pith.science/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WXZD7ITVZT2B2LRFUZMV4NR45L/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2017:WXZD7ITVZT2B2LRFUZMV4NR45L","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":"344de0b9f05f4e6980704eb16e20ae6e355b3dcc10e2d0be0ea8d85b5f7dd9d8","cross_cats_sorted":["cs.NA","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-06T23:05:10Z","title_canon_sha256":"13029903ba7673a21330464a22da8f64320910af64ac2597e2d603c4faf56ba5"},"schema_version":"1.0","source":{"id":"1711.02213","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1711.02213","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"arxiv_version","alias_value":"1711.02213v2","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1711.02213","created_at":"2026-05-18T00:29:05Z"},{"alias_kind":"pith_short_12","alias_value":"WXZD7ITVZT2B","created_at":"2026-05-18T12:31:53Z"},{"alias_kind":"pith_short_16","alias_value":"WXZD7ITVZT2B2LRF","created_at":"2026-05-18T12:31:53Z"},{"alias_kind":"pith_short_8","alias_value":"WXZD7ITV","created_at":"2026-05-18T12:31:53Z"}],"graph_snapshots":[{"event_id":"sha256:e11f917155faa6c77f584e181d8ae4db1851176cb194579d9ade99c503c7ddde","target":"graph","created_at":"2026-05-18T00:29:05Z","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":"Deep neural networks are commonly developed and trained in 32-bit floating point format. Significant gains in performance and energy efficiency could be realized by training and inference in numerical formats optimized for deep learning. Despite advances in limited precision inference in recent years, training of neural networks in low bit-width remains a challenging problem. Here we present the Flexpoint data format, aiming at a complete replacement of 32-bit floating point format training and inference, designed to support modern deep network topologies without modifications. Flexpoint tenso","authors_text":"Amir Khosrowshahi, Arjun K. Bansal, Carey Kloss, Luke Hornof, Marcel Nassar, Naveen Rao, O\\u{g}uz H. Elibol, Ruby J. Pai, Scott Gray, Stewart Hall, Tristan J. Webb, Urs K\\\"oster, William H. Constable, Xin Wang","cross_cats":["cs.NA","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-06T23:05:10Z","title":"Flexpoint: An Adaptive Numerical Format for Efficient Training of Deep Neural Networks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1711.02213","kind":"arxiv","version":2},"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":"sha256:296bc82164e720ec8c7a0d30de685015a866dffe2f2a346fc6d8d81f02818e5d","target":"record","created_at":"2026-05-18T00:29:05Z","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":"344de0b9f05f4e6980704eb16e20ae6e355b3dcc10e2d0be0ea8d85b5f7dd9d8","cross_cats_sorted":["cs.NA","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-11-06T23:05:10Z","title_canon_sha256":"13029903ba7673a21330464a22da8f64320910af64ac2597e2d603c4faf56ba5"},"schema_version":"1.0","source":{"id":"1711.02213","kind":"arxiv","version":2}},"canonical_sha256":"b5f23fa275ccf41d2e25a6595e363ceaeb2d0d1b17a2722127c3df249841ee46","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b5f23fa275ccf41d2e25a6595e363ceaeb2d0d1b17a2722127c3df249841ee46","first_computed_at":"2026-05-18T00:29:05.392173Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:29:05.392173Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8PmTteSglD6TaGqDvSfA2GQSpEN/Z5D9FoT7SxLwMhy2sIe45kcU58lset5ucseMOFEX0xt0DPdM8QS2jKX7DQ==","signature_status":"signed_v1","signed_at":"2026-05-18T00:29:05.392829Z","signed_message":"canonical_sha256_bytes"},"source_id":"1711.02213","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:296bc82164e720ec8c7a0d30de685015a866dffe2f2a346fc6d8d81f02818e5d","sha256:e11f917155faa6c77f584e181d8ae4db1851176cb194579d9ade99c503c7ddde"],"state_sha256":"0965460776a091ff6eb69d3c764bd24383ab5c37f37ab2f2d90e9ed45ca922fb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eCa47M1AGMbBBhI42jQ/1NZvcmfu1k8JsU1CzBuQ9ZlMSWm9/64H+f0imPNQIz3PXgiIfQCnKwpGIFAmT5iTAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T10:57:42.244540Z","bundle_sha256":"4894835df0f04b0b504af0780488d83d3736df439fc589b75cfcd96b1221529d"}}