{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:TJ77MKGIYVJE4LNYFYW7EA7KKL","short_pith_number":"pith:TJ77MKGI","canonical_record":{"source":{"id":"1812.00542","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-12-03T03:39:08Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"36cdfc4d2d7f08a92e4fe4b8cddca0f7051ea6ee8e78872a87b6b8555bb33e6a","abstract_canon_sha256":"e148423fa77a4e98c903f3e570d804c0a50e6ff4ec66ba4b38d4c2614b319d12"},"schema_version":"1.0"},"canonical_sha256":"9a7ff628c8c5524e2db82e2df203ea52d204c51ea1107ce7e668f38c798ded81","source":{"kind":"arxiv","id":"1812.00542","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1812.00542","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"arxiv_version","alias_value":"1812.00542v1","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.00542","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"pith_short_12","alias_value":"TJ77MKGIYVJE","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_16","alias_value":"TJ77MKGIYVJE4LNY","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_8","alias_value":"TJ77MKGI","created_at":"2026-05-18T12:32:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:TJ77MKGIYVJE4LNYFYW7EA7KKL","target":"record","payload":{"canonical_record":{"source":{"id":"1812.00542","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-12-03T03:39:08Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"36cdfc4d2d7f08a92e4fe4b8cddca0f7051ea6ee8e78872a87b6b8555bb33e6a","abstract_canon_sha256":"e148423fa77a4e98c903f3e570d804c0a50e6ff4ec66ba4b38d4c2614b319d12"},"schema_version":"1.0"},"canonical_sha256":"9a7ff628c8c5524e2db82e2df203ea52d204c51ea1107ce7e668f38c798ded81","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:59:21.054709Z","signature_b64":"Jl9HL1A+gnhJFdneHy2a2S8fDE97qhfuAJyv1Igv6XSshboqvZso0vBpYgwWnCeatvXQvZXettnlB3umbUGnAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a7ff628c8c5524e2db82e2df203ea52d204c51ea1107ce7e668f38c798ded81","last_reissued_at":"2026-05-17T23:59:21.054298Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:59:21.054298Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1812.00542","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":"2026-05-17T23:59:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Gw2e+5szgVhVm3avwzSCpc9FEDmsXBeXbWXAM774+0BjSc8byB7VFJVoq9Xx6C3WNKtPBkL5laIHzVnyGOJuBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T02:05:11.897549Z"},"content_sha256":"2673875308cee7dbcf6467ddffa01c6be28255016e5f687f824f82289bc69f01","schema_version":"1.0","event_id":"sha256:2673875308cee7dbcf6467ddffa01c6be28255016e5f687f824f82289bc69f01"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:TJ77MKGIYVJE4LNYFYW7EA7KKL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Towards Theoretical Understanding of Large Batch Training in Stochastic Gradient Descent","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Xiaowu Dai, Yuhua Zhu","submitted_at":"2018-12-03T03:39:08Z","abstract_excerpt":"Stochastic gradient descent (SGD) is almost ubiquitously used for training non-convex optimization tasks. Recently, a hypothesis proposed by Keskar et al. [2017] that large batch methods tend to converge to sharp minimizers has received increasing attention. We theoretically justify this hypothesis by providing new properties of SGD in both finite-time and asymptotic regimes. In particular, we give an explicit escaping time of SGD from a local minimum in the finite-time regime and prove that SGD tends to converge to flatter minima in the asymptotic regime (although may take exponential time to"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.00542","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":"2026-05-17T23:59:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HE1NS1LWtWicFuB7Q1V6qcVq1sJ7wRYwAkRpnJc8/+m+W62psb8qhF5ds8oU/45S/a6lrYB+JIOeY4VV5w20Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T02:05:11.898101Z"},"content_sha256":"8f694065d6dea4d103025a577fae1c4dd1e9d25ea689c36f51f9498862f5e3b9","schema_version":"1.0","event_id":"sha256:8f694065d6dea4d103025a577fae1c4dd1e9d25ea689c36f51f9498862f5e3b9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/bundle.json","state_url":"https://pith.science/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/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":"2026-06-05T02:05:11Z","links":{"resolver":"https://pith.science/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL","bundle":"https://pith.science/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/bundle.json","state":"https://pith.science/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TJ77MKGIYVJE4LNYFYW7EA7KKL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:TJ77MKGIYVJE4LNYFYW7EA7KKL","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":"e148423fa77a4e98c903f3e570d804c0a50e6ff4ec66ba4b38d4c2614b319d12","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-12-03T03:39:08Z","title_canon_sha256":"36cdfc4d2d7f08a92e4fe4b8cddca0f7051ea6ee8e78872a87b6b8555bb33e6a"},"schema_version":"1.0","source":{"id":"1812.00542","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1812.00542","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"arxiv_version","alias_value":"1812.00542v1","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.00542","created_at":"2026-05-17T23:59:21Z"},{"alias_kind":"pith_short_12","alias_value":"TJ77MKGIYVJE","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_16","alias_value":"TJ77MKGIYVJE4LNY","created_at":"2026-05-18T12:32:53Z"},{"alias_kind":"pith_short_8","alias_value":"TJ77MKGI","created_at":"2026-05-18T12:32:53Z"}],"graph_snapshots":[{"event_id":"sha256:8f694065d6dea4d103025a577fae1c4dd1e9d25ea689c36f51f9498862f5e3b9","target":"graph","created_at":"2026-05-17T23:59:21Z","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":"Stochastic gradient descent (SGD) is almost ubiquitously used for training non-convex optimization tasks. Recently, a hypothesis proposed by Keskar et al. [2017] that large batch methods tend to converge to sharp minimizers has received increasing attention. We theoretically justify this hypothesis by providing new properties of SGD in both finite-time and asymptotic regimes. In particular, we give an explicit escaping time of SGD from a local minimum in the finite-time regime and prove that SGD tends to converge to flatter minima in the asymptotic regime (although may take exponential time to","authors_text":"Xiaowu Dai, Yuhua Zhu","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-12-03T03:39:08Z","title":"Towards Theoretical Understanding of Large Batch Training in Stochastic Gradient Descent"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.00542","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":"sha256:2673875308cee7dbcf6467ddffa01c6be28255016e5f687f824f82289bc69f01","target":"record","created_at":"2026-05-17T23:59:21Z","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":"e148423fa77a4e98c903f3e570d804c0a50e6ff4ec66ba4b38d4c2614b319d12","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-12-03T03:39:08Z","title_canon_sha256":"36cdfc4d2d7f08a92e4fe4b8cddca0f7051ea6ee8e78872a87b6b8555bb33e6a"},"schema_version":"1.0","source":{"id":"1812.00542","kind":"arxiv","version":1}},"canonical_sha256":"9a7ff628c8c5524e2db82e2df203ea52d204c51ea1107ce7e668f38c798ded81","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9a7ff628c8c5524e2db82e2df203ea52d204c51ea1107ce7e668f38c798ded81","first_computed_at":"2026-05-17T23:59:21.054298Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:59:21.054298Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Jl9HL1A+gnhJFdneHy2a2S8fDE97qhfuAJyv1Igv6XSshboqvZso0vBpYgwWnCeatvXQvZXettnlB3umbUGnAw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:59:21.054709Z","signed_message":"canonical_sha256_bytes"},"source_id":"1812.00542","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2673875308cee7dbcf6467ddffa01c6be28255016e5f687f824f82289bc69f01","sha256:8f694065d6dea4d103025a577fae1c4dd1e9d25ea689c36f51f9498862f5e3b9"],"state_sha256":"df0aed172e8b7e88b76f3c172ae65cbb26e9d827699aa1f764fea7b8fbb12f19"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qAPqmyCQ0eppf5wJr11MOlp8hqWlggjVSO9ybESmqEHIwvgUA/UQ7c3KsQRYnRKf6yIzXmBHLnBvqUPGXI5VDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T02:05:11.900982Z","bundle_sha256":"28328c317c807f8794e5b2a745f7fe46f7686177003f6eb59621573b01f3a22e"}}