{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:JUV2VY6FK23SMEDNHMW23VESQV","short_pith_number":"pith:JUV2VY6F","canonical_record":{"source":{"id":"1906.10822","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-06-26T02:54:26Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"017ea2fee401be6e6725a56392c7d234aeeb379c6ec65bd767c0b25bf271c942","abstract_canon_sha256":"cb9561576a0efcfb74f2ac6093b9ac42a7843881450c9f2304f197241d9d3ec0"},"schema_version":"1.0"},"canonical_sha256":"4d2baae3c556b726106d3b2dadd492857aae386fc37e69da9f17bb140a381dd8","source":{"kind":"arxiv","id":"1906.10822","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1906.10822","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"arxiv_version","alias_value":"1906.10822v1","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1906.10822","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"pith_short_12","alias_value":"JUV2VY6FK23S","created_at":"2026-05-18T12:33:21Z"},{"alias_kind":"pith_short_16","alias_value":"JUV2VY6FK23SMEDN","created_at":"2026-05-18T12:33:21Z"},{"alias_kind":"pith_short_8","alias_value":"JUV2VY6F","created_at":"2026-05-18T12:33:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:JUV2VY6FK23SMEDNHMW23VESQV","target":"record","payload":{"canonical_record":{"source":{"id":"1906.10822","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-06-26T02:54:26Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"017ea2fee401be6e6725a56392c7d234aeeb379c6ec65bd767c0b25bf271c942","abstract_canon_sha256":"cb9561576a0efcfb74f2ac6093b9ac42a7843881450c9f2304f197241d9d3ec0"},"schema_version":"1.0"},"canonical_sha256":"4d2baae3c556b726106d3b2dadd492857aae386fc37e69da9f17bb140a381dd8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:42:13.157555Z","signature_b64":"2IP3uFo85BTBNxflZ0oazP8GkK45BEcEqe7vqf3bC7eCXevLOMJKthPPqFzkzye7ivk0TGfgRo82FM1Ursu6Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d2baae3c556b726106d3b2dadd492857aae386fc37e69da9f17bb140a381dd8","last_reissued_at":"2026-05-17T23:42:13.156874Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:42:13.156874Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1906.10822","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":"2026-05-17T23:42:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OedM+qBHq1HRK4YeZEgtUbS1nmrUTWJANP1r4CSDY88Q/d+NFogi3CRdd5yYpkPrebM2uFVBAHWKOroAkeBcDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T18:12:40.773358Z"},"content_sha256":"fa8f26c3553898e2c7d4dd198d0babcf225970e7542393959e28105a80b1a18d","schema_version":"1.0","event_id":"sha256:fa8f26c3553898e2c7d4dd198d0babcf225970e7542393959e28105a80b1a18d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:JUV2VY6FK23SMEDNHMW23VESQV","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Gradient Noise Convolution (GNC): Smoothing Loss Function for Distributed Large-Batch SGD","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Kosuke Haruki, Masahiro Ozawa, Mitsuhiro Kimura, Ryuji Sakai, Taiji Suzuki, Takeshi Toda, Yohei Hamakawa","submitted_at":"2019-06-26T02:54:26Z","abstract_excerpt":"Large-batch stochastic gradient descent (SGD) is widely used for training in distributed deep learning because of its training-time efficiency, however, extremely large-batch SGD leads to poor generalization and easily converges to sharp minima, which prevents naive large-scale data-parallel SGD (DP-SGD) from converging to good minima. To overcome this difficulty, we propose gradient noise convolution (GNC), which effectively smooths sharper minima of the loss function. For DP-SGD, GNC utilizes so-called gradient noise, which is induced by stochastic gradient variation and convolved to the los"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1906.10822","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":"2026-05-17T23:42:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"k4PtuOtXlDkJTe3JulGlKK8jWQHYbq7tA9ncGlZvhCPHTFMEe9xQbKEX80ZvjSobHpXWVEbVUbMVDJ/OJaibDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-26T18:12:40.773995Z"},"content_sha256":"55e441a4f7bbdab36e810788df2a88ae4e07b0a30156108d8c19ac012a203b33","schema_version":"1.0","event_id":"sha256:55e441a4f7bbdab36e810788df2a88ae4e07b0a30156108d8c19ac012a203b33"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JUV2VY6FK23SMEDNHMW23VESQV/bundle.json","state_url":"https://pith.science/pith/JUV2VY6FK23SMEDNHMW23VESQV/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JUV2VY6FK23SMEDNHMW23VESQV/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":"2026-05-26T18:12:40Z","links":{"resolver":"https://pith.science/pith/JUV2VY6FK23SMEDNHMW23VESQV","bundle":"https://pith.science/pith/JUV2VY6FK23SMEDNHMW23VESQV/bundle.json","state":"https://pith.science/pith/JUV2VY6FK23SMEDNHMW23VESQV/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JUV2VY6FK23SMEDNHMW23VESQV/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:JUV2VY6FK23SMEDNHMW23VESQV","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":"cb9561576a0efcfb74f2ac6093b9ac42a7843881450c9f2304f197241d9d3ec0","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-06-26T02:54:26Z","title_canon_sha256":"017ea2fee401be6e6725a56392c7d234aeeb379c6ec65bd767c0b25bf271c942"},"schema_version":"1.0","source":{"id":"1906.10822","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1906.10822","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"arxiv_version","alias_value":"1906.10822v1","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1906.10822","created_at":"2026-05-17T23:42:13Z"},{"alias_kind":"pith_short_12","alias_value":"JUV2VY6FK23S","created_at":"2026-05-18T12:33:21Z"},{"alias_kind":"pith_short_16","alias_value":"JUV2VY6FK23SMEDN","created_at":"2026-05-18T12:33:21Z"},{"alias_kind":"pith_short_8","alias_value":"JUV2VY6F","created_at":"2026-05-18T12:33:21Z"}],"graph_snapshots":[{"event_id":"sha256:55e441a4f7bbdab36e810788df2a88ae4e07b0a30156108d8c19ac012a203b33","target":"graph","created_at":"2026-05-17T23:42:13Z","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":"Large-batch stochastic gradient descent (SGD) is widely used for training in distributed deep learning because of its training-time efficiency, however, extremely large-batch SGD leads to poor generalization and easily converges to sharp minima, which prevents naive large-scale data-parallel SGD (DP-SGD) from converging to good minima. To overcome this difficulty, we propose gradient noise convolution (GNC), which effectively smooths sharper minima of the loss function. For DP-SGD, GNC utilizes so-called gradient noise, which is induced by stochastic gradient variation and convolved to the los","authors_text":"Kosuke Haruki, Masahiro Ozawa, Mitsuhiro Kimura, Ryuji Sakai, Taiji Suzuki, Takeshi Toda, Yohei Hamakawa","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-06-26T02:54:26Z","title":"Gradient Noise Convolution (GNC): Smoothing Loss Function for Distributed Large-Batch SGD"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1906.10822","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:fa8f26c3553898e2c7d4dd198d0babcf225970e7542393959e28105a80b1a18d","target":"record","created_at":"2026-05-17T23:42:13Z","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":"cb9561576a0efcfb74f2ac6093b9ac42a7843881450c9f2304f197241d9d3ec0","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-06-26T02:54:26Z","title_canon_sha256":"017ea2fee401be6e6725a56392c7d234aeeb379c6ec65bd767c0b25bf271c942"},"schema_version":"1.0","source":{"id":"1906.10822","kind":"arxiv","version":1}},"canonical_sha256":"4d2baae3c556b726106d3b2dadd492857aae386fc37e69da9f17bb140a381dd8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4d2baae3c556b726106d3b2dadd492857aae386fc37e69da9f17bb140a381dd8","first_computed_at":"2026-05-17T23:42:13.156874Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:42:13.156874Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2IP3uFo85BTBNxflZ0oazP8GkK45BEcEqe7vqf3bC7eCXevLOMJKthPPqFzkzye7ivk0TGfgRo82FM1Ursu6Cw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:42:13.157555Z","signed_message":"canonical_sha256_bytes"},"source_id":"1906.10822","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fa8f26c3553898e2c7d4dd198d0babcf225970e7542393959e28105a80b1a18d","sha256:55e441a4f7bbdab36e810788df2a88ae4e07b0a30156108d8c19ac012a203b33"],"state_sha256":"8b1413d822b52ca3204b03c4af47de35a9307c45dccc20fe31d00266cc85754a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Fh0zrTcTzz+3qmktjUQtcbn//CcEVt243b82B9VB7ftrSWBMmfW383YQFR4lxE0TfWZM3RX6H5Rlx4spUIk8Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-26T18:12:40.777408Z","bundle_sha256":"eea738a57ccc913c1e0d421a28253651ff75e4c220c72fa7aea906cc1e63a60c"}}