{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:SZQIXLNZVGTWPCWH4C4GLVHIEF","short_pith_number":"pith:SZQIXLNZ","canonical_record":{"source":{"id":"1802.06058","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-16T18:15:33Z","cross_cats_sorted":[],"title_canon_sha256":"5385086a977937b1fd47bc4e721d5e5b739c1102f279c4ead5bcc6d7d425efcc","abstract_canon_sha256":"0a98d42048021085d1940c65dc9945304e1459a4bb315083022ab7083b196bdc"},"schema_version":"1.0"},"canonical_sha256":"96608badb9a9a7678ac7e0b865d4e821599b04426da94f1426f11663ae28d3d3","source":{"kind":"arxiv","id":"1802.06058","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1802.06058","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"arxiv_version","alias_value":"1802.06058v2","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.06058","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"pith_short_12","alias_value":"SZQIXLNZVGTW","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_16","alias_value":"SZQIXLNZVGTWPCWH","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_8","alias_value":"SZQIXLNZ","created_at":"2026-05-18T12:32:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:SZQIXLNZVGTWPCWH4C4GLVHIEF","target":"record","payload":{"canonical_record":{"source":{"id":"1802.06058","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-16T18:15:33Z","cross_cats_sorted":[],"title_canon_sha256":"5385086a977937b1fd47bc4e721d5e5b739c1102f279c4ead5bcc6d7d425efcc","abstract_canon_sha256":"0a98d42048021085d1940c65dc9945304e1459a4bb315083022ab7083b196bdc"},"schema_version":"1.0"},"canonical_sha256":"96608badb9a9a7678ac7e0b865d4e821599b04426da94f1426f11663ae28d3d3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:22:54.391059Z","signature_b64":"4vV3PnCz6Jlq74Dwc+zf81fLb2K0n5YfX2FgfnLrq1IlZnEQ1lB7AnSwzGQlBzU5H3CzSph8xA606/9o7XVCBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"96608badb9a9a7678ac7e0b865d4e821599b04426da94f1426f11663ae28d3d3","last_reissued_at":"2026-05-18T00:22:54.390038Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:22:54.390038Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1802.06058","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":"2026-05-18T00:22:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oO9TvJKlTBTPoTr1PZrvAfF0ogfo1k21BDLvWLeH6C3sxbs608GSOWNFrkU6j3AkobX90XCvEPxOvsRnh/vNCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T13:34:33.524084Z"},"content_sha256":"8bdaf32b1f9a3ed1f64602e5dc1468c347f432cf57c68c0f8695aae285bf52ca","schema_version":"1.0","event_id":"sha256:8bdaf32b1f9a3ed1f64602e5dc1468c347f432cf57c68c0f8695aae285bf52ca"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:SZQIXLNZVGTWPCWH4C4GLVHIEF","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Variance-based Gradient Compression for Efficient Distributed Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Hiroto Imachi, Takuya Akiba, Yusuke Tsuzuku","submitted_at":"2018-02-16T18:15:33Z","abstract_excerpt":"Due to the substantial computational cost, training state-of-the-art deep neural networks for large-scale datasets often requires distributed training using multiple computation workers. However, by nature, workers need to frequently communicate gradients, causing severe bottlenecks, especially on lower bandwidth connections. A few methods have been proposed to compress gradient for efficient communication, but they either suffer a low compression ratio or significantly harm the resulting model accuracy, particularly when applied to convolutional neural networks. To address these issues, we pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.06058","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":"2026-05-18T00:22:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hQxRKWy/571Zxzyj2V4EPe00VDQh61/4WC5U8pdfWWKASQGO2zykR/9xgUPAZ8iwBZEUWhGYTyjzu8Db5lRBDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-28T13:34:33.524456Z"},"content_sha256":"d2ebcc3c9992d0c8954620a0a9be25b7fe9a41d4a62fc53648f5e1cbec457313","schema_version":"1.0","event_id":"sha256:d2ebcc3c9992d0c8954620a0a9be25b7fe9a41d4a62fc53648f5e1cbec457313"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/bundle.json","state_url":"https://pith.science/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/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":"2026-05-28T13:34:33Z","links":{"resolver":"https://pith.science/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF","bundle":"https://pith.science/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/bundle.json","state":"https://pith.science/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SZQIXLNZVGTWPCWH4C4GLVHIEF/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:SZQIXLNZVGTWPCWH4C4GLVHIEF","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":"0a98d42048021085d1940c65dc9945304e1459a4bb315083022ab7083b196bdc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-16T18:15:33Z","title_canon_sha256":"5385086a977937b1fd47bc4e721d5e5b739c1102f279c4ead5bcc6d7d425efcc"},"schema_version":"1.0","source":{"id":"1802.06058","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1802.06058","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"arxiv_version","alias_value":"1802.06058v2","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.06058","created_at":"2026-05-18T00:22:54Z"},{"alias_kind":"pith_short_12","alias_value":"SZQIXLNZVGTW","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_16","alias_value":"SZQIXLNZVGTWPCWH","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_8","alias_value":"SZQIXLNZ","created_at":"2026-05-18T12:32:53Z"}],"graph_snapshots":[{"event_id":"sha256:d2ebcc3c9992d0c8954620a0a9be25b7fe9a41d4a62fc53648f5e1cbec457313","target":"graph","created_at":"2026-05-18T00:22:54Z","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":"Due to the substantial computational cost, training state-of-the-art deep neural networks for large-scale datasets often requires distributed training using multiple computation workers. However, by nature, workers need to frequently communicate gradients, causing severe bottlenecks, especially on lower bandwidth connections. A few methods have been proposed to compress gradient for efficient communication, but they either suffer a low compression ratio or significantly harm the resulting model accuracy, particularly when applied to convolutional neural networks. To address these issues, we pr","authors_text":"Hiroto Imachi, Takuya Akiba, Yusuke Tsuzuku","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-16T18:15:33Z","title":"Variance-based Gradient Compression for Efficient Distributed Deep Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.06058","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":"sha256:8bdaf32b1f9a3ed1f64602e5dc1468c347f432cf57c68c0f8695aae285bf52ca","target":"record","created_at":"2026-05-18T00:22:54Z","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":"0a98d42048021085d1940c65dc9945304e1459a4bb315083022ab7083b196bdc","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-16T18:15:33Z","title_canon_sha256":"5385086a977937b1fd47bc4e721d5e5b739c1102f279c4ead5bcc6d7d425efcc"},"schema_version":"1.0","source":{"id":"1802.06058","kind":"arxiv","version":2}},"canonical_sha256":"96608badb9a9a7678ac7e0b865d4e821599b04426da94f1426f11663ae28d3d3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"96608badb9a9a7678ac7e0b865d4e821599b04426da94f1426f11663ae28d3d3","first_computed_at":"2026-05-18T00:22:54.390038Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:22:54.390038Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4vV3PnCz6Jlq74Dwc+zf81fLb2K0n5YfX2FgfnLrq1IlZnEQ1lB7AnSwzGQlBzU5H3CzSph8xA606/9o7XVCBw==","signature_status":"signed_v1","signed_at":"2026-05-18T00:22:54.391059Z","signed_message":"canonical_sha256_bytes"},"source_id":"1802.06058","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8bdaf32b1f9a3ed1f64602e5dc1468c347f432cf57c68c0f8695aae285bf52ca","sha256:d2ebcc3c9992d0c8954620a0a9be25b7fe9a41d4a62fc53648f5e1cbec457313"],"state_sha256":"77952d88279952a12279f25069df13caef2b77e41086b1c4dda1898326e7723d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"d+qk+VCWoKfb24dVrUahgYJT4CZjnKU6zYC7eKdDVkyJDz4CtRT3+TJPCa5P5GymU4oCjegjJd4CE82MbPRpAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-28T13:34:33.526471Z","bundle_sha256":"fb6e6e1d5db828918b8bec371659f1845236b8edf8d2cdd3988594fc87e2114a"}}