{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:MKNUG7SUKMBS6FMZAGXOGVGRTR","short_pith_number":"pith:MKNUG7SU","canonical_record":{"source":{"id":"1908.05460","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-08-15T09:11:31Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"df83a4cdde0715f37d65e24a856434a80272f919ef1e5b732b70bd37c7ffa956","abstract_canon_sha256":"ea14958606df333d9e8b64aadeeb5d02a13b754c283eaf5db10520ad164bd413"},"schema_version":"1.0"},"canonical_sha256":"629b437e5453032f159901aee354d19c72b78304b6f8d971596e76817af647e6","source":{"kind":"arxiv","id":"1908.05460","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1908.05460","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"arxiv_version","alias_value":"1908.05460v1","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.05460","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_12","alias_value":"MKNUG7SUKMBS","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_16","alias_value":"MKNUG7SUKMBS6FMZ","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_8","alias_value":"MKNUG7SU","created_at":"2026-07-04T23:56:52Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:MKNUG7SUKMBS6FMZAGXOGVGRTR","target":"record","payload":{"canonical_record":{"source":{"id":"1908.05460","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-08-15T09:11:31Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"df83a4cdde0715f37d65e24a856434a80272f919ef1e5b732b70bd37c7ffa956","abstract_canon_sha256":"ea14958606df333d9e8b64aadeeb5d02a13b754c283eaf5db10520ad164bd413"},"schema_version":"1.0"},"canonical_sha256":"629b437e5453032f159901aee354d19c72b78304b6f8d971596e76817af647e6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-04T23:56:52.547421Z","signature_b64":"JX0/ZmfTJxB7MFOBBMDyT388SL1KpJwc6dD2rm15H+Wq/rfjOnwVGTTu9F+wLkkFHhTpkMRoI89TNVXAL82XBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"629b437e5453032f159901aee354d19c72b78304b6f8d971596e76817af647e6","last_reissued_at":"2026-07-04T23:56:52.546956Z","signature_status":"signed_v1","first_computed_at":"2026-07-04T23:56:52.546956Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1908.05460","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":"2026-07-04T23:56:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OQJH8WTXQ3Sc1qeqa0ujWVRfXXJtjD42KccIvwurG2Pm9Ef+SBKoHaaq9L66yPIBoD+Z3mXDiYU1eEE4edveAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T01:24:52.389708Z"},"content_sha256":"a946a47b52d97e29caa16c6082f12ff71632ff25465793a4c73dd8511de999da","schema_version":"1.0","event_id":"sha256:a946a47b52d97e29caa16c6082f12ff71632ff25465793a4c73dd8511de999da"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:MKNUG7SUKMBS6FMZAGXOGVGRTR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Accelerated CNN Training Through Gradient Approximation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Sree Harsha Nelaturu, Ziheng Wang","submitted_at":"2019-08-15T09:11:31Z","abstract_excerpt":"Training deep convolutional neural networks such as VGG and ResNet by gradient descent is an expensive exercise requiring specialized hardware such as GPUs. Recent works have examined the possibility of approximating the gradient computation while maintaining the same convergence properties. While promising, the approximations only work on relatively small datasets such as MNIST. They also fail to achieve real wall-clock speedups due to lack of efficient GPU implementations of the proposed approximation methods. In this work, we explore three alternative methods to approximate gradients, with "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.05460","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1908.05460/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"2026-07-04T23:56:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P8HIsJPoUGpsABahaZaLAVUK3Z9HEjT2sm5OpYnyeifBkOM4vl0IsFvL2Z2RN6Gd4WeybsBkGAa+V21gvx6hBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T01:24:52.390235Z"},"content_sha256":"8bd4076ea9d39077bf14223e7a7834b6ad31273736dd21b9a47aa9bb1f1652ca","schema_version":"1.0","event_id":"sha256:8bd4076ea9d39077bf14223e7a7834b6ad31273736dd21b9a47aa9bb1f1652ca"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/bundle.json","state_url":"https://pith.science/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/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":"2026-08-21T01:24:52Z","links":{"resolver":"https://pith.science/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR","bundle":"https://pith.science/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/bundle.json","state":"https://pith.science/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MKNUG7SUKMBS6FMZAGXOGVGRTR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:MKNUG7SUKMBS6FMZAGXOGVGRTR","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":"ea14958606df333d9e8b64aadeeb5d02a13b754c283eaf5db10520ad164bd413","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-08-15T09:11:31Z","title_canon_sha256":"df83a4cdde0715f37d65e24a856434a80272f919ef1e5b732b70bd37c7ffa956"},"schema_version":"1.0","source":{"id":"1908.05460","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1908.05460","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"arxiv_version","alias_value":"1908.05460v1","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1908.05460","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_12","alias_value":"MKNUG7SUKMBS","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_16","alias_value":"MKNUG7SUKMBS6FMZ","created_at":"2026-07-04T23:56:52Z"},{"alias_kind":"pith_short_8","alias_value":"MKNUG7SU","created_at":"2026-07-04T23:56:52Z"}],"graph_snapshots":[{"event_id":"sha256:8bd4076ea9d39077bf14223e7a7834b6ad31273736dd21b9a47aa9bb1f1652ca","target":"graph","created_at":"2026-07-04T23:56:52Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1908.05460/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Training deep convolutional neural networks such as VGG and ResNet by gradient descent is an expensive exercise requiring specialized hardware such as GPUs. Recent works have examined the possibility of approximating the gradient computation while maintaining the same convergence properties. While promising, the approximations only work on relatively small datasets such as MNIST. They also fail to achieve real wall-clock speedups due to lack of efficient GPU implementations of the proposed approximation methods. In this work, we explore three alternative methods to approximate gradients, with ","authors_text":"Sree Harsha Nelaturu, Ziheng Wang","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-08-15T09:11:31Z","title":"Accelerated CNN Training Through Gradient Approximation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1908.05460","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:a946a47b52d97e29caa16c6082f12ff71632ff25465793a4c73dd8511de999da","target":"record","created_at":"2026-07-04T23:56:52Z","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":"ea14958606df333d9e8b64aadeeb5d02a13b754c283eaf5db10520ad164bd413","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-08-15T09:11:31Z","title_canon_sha256":"df83a4cdde0715f37d65e24a856434a80272f919ef1e5b732b70bd37c7ffa956"},"schema_version":"1.0","source":{"id":"1908.05460","kind":"arxiv","version":1}},"canonical_sha256":"629b437e5453032f159901aee354d19c72b78304b6f8d971596e76817af647e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"629b437e5453032f159901aee354d19c72b78304b6f8d971596e76817af647e6","first_computed_at":"2026-07-04T23:56:52.546956Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-04T23:56:52.546956Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JX0/ZmfTJxB7MFOBBMDyT388SL1KpJwc6dD2rm15H+Wq/rfjOnwVGTTu9F+wLkkFHhTpkMRoI89TNVXAL82XBw==","signature_status":"signed_v1","signed_at":"2026-07-04T23:56:52.547421Z","signed_message":"canonical_sha256_bytes"},"source_id":"1908.05460","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a946a47b52d97e29caa16c6082f12ff71632ff25465793a4c73dd8511de999da","sha256:8bd4076ea9d39077bf14223e7a7834b6ad31273736dd21b9a47aa9bb1f1652ca"],"state_sha256":"22316656cd1ff0fafee08dc391472370ee60f6fa0a021b06f3558705260ec85e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QrMjSupac6gfI4nUhMqP4A+1UrTOxm9uNiixSnfC20basa7Ux7Yd3S3HVJ5ejcy/ERgWFaWqaFPTSlPYtze+Cw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-21T01:24:52.394645Z","bundle_sha256":"0b0336818b3b9ba266b058c83f574b048e01bfcd85b37f3a2bd1a7a6808fbf35"}}