{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:2RARNKYQJ2OENEMZQNP5RY3FOP","short_pith_number":"pith:2RARNKYQ","canonical_record":{"source":{"id":"1806.04854","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-13T05:45:22Z","cross_cats_sorted":["cs.AI","cs.LG","stat.CO"],"title_canon_sha256":"12cbe9a18e8b793a67ba0223e35e72539881bc2d02a483d3282c567352d62f67","abstract_canon_sha256":"34fb5aca48faa5df0222220366dd4e8155e95faee75724718cf5ba40183cf670"},"schema_version":"1.0"},"canonical_sha256":"d44116ab104e9c469199835fd8e36573ef388458ab786a18d56a7b2b3f462118","source":{"kind":"arxiv","id":"1806.04854","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1806.04854","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"arxiv_version","alias_value":"1806.04854v3","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1806.04854","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"pith_short_12","alias_value":"2RARNKYQJ2OE","created_at":"2026-05-18T12:32:02Z"},{"alias_kind":"pith_short_16","alias_value":"2RARNKYQJ2OENEMZ","created_at":"2026-05-18T12:32:02Z"},{"alias_kind":"pith_short_8","alias_value":"2RARNKYQ","created_at":"2026-05-18T12:32:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:2RARNKYQJ2OENEMZQNP5RY3FOP","target":"record","payload":{"canonical_record":{"source":{"id":"1806.04854","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-13T05:45:22Z","cross_cats_sorted":["cs.AI","cs.LG","stat.CO"],"title_canon_sha256":"12cbe9a18e8b793a67ba0223e35e72539881bc2d02a483d3282c567352d62f67","abstract_canon_sha256":"34fb5aca48faa5df0222220366dd4e8155e95faee75724718cf5ba40183cf670"},"schema_version":"1.0"},"canonical_sha256":"d44116ab104e9c469199835fd8e36573ef388458ab786a18d56a7b2b3f462118","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:09:03.757971Z","signature_b64":"vnnx+ilPYVh+yvCgHbpWR9EJxYeoBSFWULTCORud1FW3BeYEqPNUJ22LRBBSE25qb+00SEnMenkbOa7aAZ1WAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d44116ab104e9c469199835fd8e36573ef388458ab786a18d56a7b2b3f462118","last_reissued_at":"2026-05-18T00:09:03.757435Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:09:03.757435Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1806.04854","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T00:09:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dbO3L8HIVbPedSdXyTsz5boDb0k4YkGAxbEcvT3jxXNDuFX7jg6QOuPHXYtkDILWLqGEHsy8o+8mHVtF+ct4DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T14:56:23.942067Z"},"content_sha256":"d823dceaeecfbf75a64697b17a15d7b237f76db8818cb74801373700ca1eba6c","schema_version":"1.0","event_id":"sha256:d823dceaeecfbf75a64697b17a15d7b237f76db8818cb74801373700ca1eba6c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:2RARNKYQJ2OENEMZQNP5RY3FOP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fast and Scalable Bayesian Deep Learning by Weight-Perturbation in Adam","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","stat.CO"],"primary_cat":"stat.ML","authors_text":"Akash Srivastava, Didrik Nielsen, Mohammad Emtiyaz Khan, Voot Tangkaratt, Wu Lin, Yarin Gal","submitted_at":"2018-06-13T05:45:22Z","abstract_excerpt":"Uncertainty computation in deep learning is essential to design robust and reliable systems. Variational inference (VI) is a promising approach for such computation, but requires more effort to implement and execute compared to maximum-likelihood methods. In this paper, we propose new natural-gradient algorithms to reduce such efforts for Gaussian mean-field VI. Our algorithms can be implemented within the Adam optimizer by perturbing the network weights during gradient evaluations, and uncertainty estimates can be cheaply obtained by using the vector that adapts the learning rate. This requir"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1806.04854","kind":"arxiv","version":3},"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":"2026-05-18T00:09:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+E0OBd7AlMTid4q+qhOti7q9v5Gpamn7uvT575hHFsSxDpPXCaBXxtGMc28lm6aotU4Vb6BTUG2fysP+WW7cDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-29T14:56:23.942443Z"},"content_sha256":"52a399dee31154c155244bb785802444e4caf7a95dfa99607823852a14b49bfa","schema_version":"1.0","event_id":"sha256:52a399dee31154c155244bb785802444e4caf7a95dfa99607823852a14b49bfa"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/bundle.json","state_url":"https://pith.science/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/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":"2026-05-29T14:56:23Z","links":{"resolver":"https://pith.science/pith/2RARNKYQJ2OENEMZQNP5RY3FOP","bundle":"https://pith.science/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/bundle.json","state":"https://pith.science/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2RARNKYQJ2OENEMZQNP5RY3FOP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:2RARNKYQJ2OENEMZQNP5RY3FOP","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":"34fb5aca48faa5df0222220366dd4e8155e95faee75724718cf5ba40183cf670","cross_cats_sorted":["cs.AI","cs.LG","stat.CO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-13T05:45:22Z","title_canon_sha256":"12cbe9a18e8b793a67ba0223e35e72539881bc2d02a483d3282c567352d62f67"},"schema_version":"1.0","source":{"id":"1806.04854","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1806.04854","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"arxiv_version","alias_value":"1806.04854v3","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1806.04854","created_at":"2026-05-18T00:09:03Z"},{"alias_kind":"pith_short_12","alias_value":"2RARNKYQJ2OE","created_at":"2026-05-18T12:32:02Z"},{"alias_kind":"pith_short_16","alias_value":"2RARNKYQJ2OENEMZ","created_at":"2026-05-18T12:32:02Z"},{"alias_kind":"pith_short_8","alias_value":"2RARNKYQ","created_at":"2026-05-18T12:32:02Z"}],"graph_snapshots":[{"event_id":"sha256:52a399dee31154c155244bb785802444e4caf7a95dfa99607823852a14b49bfa","target":"graph","created_at":"2026-05-18T00:09:03Z","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":"Uncertainty computation in deep learning is essential to design robust and reliable systems. Variational inference (VI) is a promising approach for such computation, but requires more effort to implement and execute compared to maximum-likelihood methods. In this paper, we propose new natural-gradient algorithms to reduce such efforts for Gaussian mean-field VI. Our algorithms can be implemented within the Adam optimizer by perturbing the network weights during gradient evaluations, and uncertainty estimates can be cheaply obtained by using the vector that adapts the learning rate. This requir","authors_text":"Akash Srivastava, Didrik Nielsen, Mohammad Emtiyaz Khan, Voot Tangkaratt, Wu Lin, Yarin Gal","cross_cats":["cs.AI","cs.LG","stat.CO"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-13T05:45:22Z","title":"Fast and Scalable Bayesian Deep Learning by Weight-Perturbation in Adam"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1806.04854","kind":"arxiv","version":3},"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:d823dceaeecfbf75a64697b17a15d7b237f76db8818cb74801373700ca1eba6c","target":"record","created_at":"2026-05-18T00:09:03Z","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":"34fb5aca48faa5df0222220366dd4e8155e95faee75724718cf5ba40183cf670","cross_cats_sorted":["cs.AI","cs.LG","stat.CO"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-13T05:45:22Z","title_canon_sha256":"12cbe9a18e8b793a67ba0223e35e72539881bc2d02a483d3282c567352d62f67"},"schema_version":"1.0","source":{"id":"1806.04854","kind":"arxiv","version":3}},"canonical_sha256":"d44116ab104e9c469199835fd8e36573ef388458ab786a18d56a7b2b3f462118","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d44116ab104e9c469199835fd8e36573ef388458ab786a18d56a7b2b3f462118","first_computed_at":"2026-05-18T00:09:03.757435Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T00:09:03.757435Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vnnx+ilPYVh+yvCgHbpWR9EJxYeoBSFWULTCORud1FW3BeYEqPNUJ22LRBBSE25qb+00SEnMenkbOa7aAZ1WAA==","signature_status":"signed_v1","signed_at":"2026-05-18T00:09:03.757971Z","signed_message":"canonical_sha256_bytes"},"source_id":"1806.04854","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d823dceaeecfbf75a64697b17a15d7b237f76db8818cb74801373700ca1eba6c","sha256:52a399dee31154c155244bb785802444e4caf7a95dfa99607823852a14b49bfa"],"state_sha256":"dbd26ff319bfd847d568c857d428547ef6727b9c3fc4a8522bac2b64f57380d8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"i9gYRxl2GufeiTv2EW1EYdsgZidzbLkX0orFJQ28Q0Y7iVvsaw2z1vt+uvZbXLmQFEh9ZDQXkJ7KmczYEo06Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-29T14:56:23.944466Z","bundle_sha256":"e421005d85a719bbe74bfacc28d47f5197bdd6bde07e9cb96a4581cdb613c862"}}