{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:VVGZHB4TBAJGKVG7FZLFAFPLPG","short_pith_number":"pith:VVGZHB4T","canonical_record":{"source":{"id":"2306.03968","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-06T19:02:57Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a0ea1d92f0b6e3ad7ec98fb5ef484128292e7673284cc057aa73c0f8f8db0efd","abstract_canon_sha256":"975d1e5093236e8dd1ef52d231a1ffe01dcc28e02b11411eef7e8de2bb122a39"},"schema_version":"1.0"},"canonical_sha256":"ad4d93879308126554df2e565015eb799f62fbfbfcfba5e5b69237fb55a78509","source":{"kind":"arxiv","id":"2306.03968","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.03968","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"arxiv_version","alias_value":"2306.03968v1","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.03968","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_12","alias_value":"VVGZHB4TBAJG","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_16","alias_value":"VVGZHB4TBAJGKVG7","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_8","alias_value":"VVGZHB4T","created_at":"2026-07-05T06:18:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:VVGZHB4TBAJGKVG7FZLFAFPLPG","target":"record","payload":{"canonical_record":{"source":{"id":"2306.03968","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-06T19:02:57Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a0ea1d92f0b6e3ad7ec98fb5ef484128292e7673284cc057aa73c0f8f8db0efd","abstract_canon_sha256":"975d1e5093236e8dd1ef52d231a1ffe01dcc28e02b11411eef7e8de2bb122a39"},"schema_version":"1.0"},"canonical_sha256":"ad4d93879308126554df2e565015eb799f62fbfbfcfba5e5b69237fb55a78509","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:18:24.409923Z","signature_b64":"kt3K7axsY85TNnwdy+uqZh8x22OjmtI6PQZ13dRW/qKYocqvWTMTitWFMIdiUNACSvjg2zROFXE8dddOXRZLCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad4d93879308126554df2e565015eb799f62fbfbfcfba5e5b69237fb55a78509","last_reissued_at":"2026-07-05T06:18:24.409468Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:18:24.409468Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2306.03968","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":"2026-07-05T06:18:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"u4oy4LwgUTGjR0WlM5wzGyKZDe1F1G0p+8fPXozCvdZRF2CipqQ9FoFAS+ZxWGYZ900clMqXqODAtwnD7jDPCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T23:09:02.154254Z"},"content_sha256":"d01d93e1e4fd019905f20349e9cd0562620e3790edf9ce29eee85ab01914883b","schema_version":"1.0","event_id":"sha256:d01d93e1e4fd019905f20349e9cd0562620e3790edf9ce29eee85ab01914883b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:VVGZHB4TBAJGKVG7FZLFAFPLPG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Stochastic Marginal Likelihood Gradients using Neural Tangent Kernels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Alexander Immer, Bernhard Sch\\\"olkopf, Gunnar R\\\"atsch, Mark van der Wilk, Tycho F. A. van der Ouderaa","submitted_at":"2023-06-06T19:02:57Z","abstract_excerpt":"Selecting hyperparameters in deep learning greatly impacts its effectiveness but requires manual effort and expertise. Recent works show that Bayesian model selection with Laplace approximations can allow to optimize such hyperparameters just like standard neural network parameters using gradients and on the training data. However, estimating a single hyperparameter gradient requires a pass through the entire dataset, limiting the scalability of such algorithms. In this work, we overcome this issue by introducing lower bounds to the linearized Laplace approximation of the marginal likelihood. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.03968","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/2306.03968/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":"2026-07-05T06:18:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"g7O0OeHjIVcPOEAZMorcY8M6Kr2zv0UnbD6ubQbHPXs27/RD8KJwnqYZaIn9CeDCKl+cPAmS8bHPqI/1fq/KAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T23:09:02.154643Z"},"content_sha256":"e3bb103edbcc715a987ffd67576c8c1b3e47c5ddfd4bc83de0b73205c036ca6f","schema_version":"1.0","event_id":"sha256:e3bb103edbcc715a987ffd67576c8c1b3e47c5ddfd4bc83de0b73205c036ca6f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/bundle.json","state_url":"https://pith.science/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/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":"2026-07-25T23:09:02Z","links":{"resolver":"https://pith.science/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG","bundle":"https://pith.science/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/bundle.json","state":"https://pith.science/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VVGZHB4TBAJGKVG7FZLFAFPLPG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:VVGZHB4TBAJGKVG7FZLFAFPLPG","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":"975d1e5093236e8dd1ef52d231a1ffe01dcc28e02b11411eef7e8de2bb122a39","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-06T19:02:57Z","title_canon_sha256":"a0ea1d92f0b6e3ad7ec98fb5ef484128292e7673284cc057aa73c0f8f8db0efd"},"schema_version":"1.0","source":{"id":"2306.03968","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.03968","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"arxiv_version","alias_value":"2306.03968v1","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.03968","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_12","alias_value":"VVGZHB4TBAJG","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_16","alias_value":"VVGZHB4TBAJGKVG7","created_at":"2026-07-05T06:18:24Z"},{"alias_kind":"pith_short_8","alias_value":"VVGZHB4T","created_at":"2026-07-05T06:18:24Z"}],"graph_snapshots":[{"event_id":"sha256:e3bb103edbcc715a987ffd67576c8c1b3e47c5ddfd4bc83de0b73205c036ca6f","target":"graph","created_at":"2026-07-05T06:18:24Z","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/2306.03968/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Selecting hyperparameters in deep learning greatly impacts its effectiveness but requires manual effort and expertise. Recent works show that Bayesian model selection with Laplace approximations can allow to optimize such hyperparameters just like standard neural network parameters using gradients and on the training data. However, estimating a single hyperparameter gradient requires a pass through the entire dataset, limiting the scalability of such algorithms. In this work, we overcome this issue by introducing lower bounds to the linearized Laplace approximation of the marginal likelihood. ","authors_text":"Alexander Immer, Bernhard Sch\\\"olkopf, Gunnar R\\\"atsch, Mark van der Wilk, Tycho F. A. van der Ouderaa","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-06T19:02:57Z","title":"Stochastic Marginal Likelihood Gradients using Neural Tangent Kernels"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.03968","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:d01d93e1e4fd019905f20349e9cd0562620e3790edf9ce29eee85ab01914883b","target":"record","created_at":"2026-07-05T06:18:24Z","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":"975d1e5093236e8dd1ef52d231a1ffe01dcc28e02b11411eef7e8de2bb122a39","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-06-06T19:02:57Z","title_canon_sha256":"a0ea1d92f0b6e3ad7ec98fb5ef484128292e7673284cc057aa73c0f8f8db0efd"},"schema_version":"1.0","source":{"id":"2306.03968","kind":"arxiv","version":1}},"canonical_sha256":"ad4d93879308126554df2e565015eb799f62fbfbfcfba5e5b69237fb55a78509","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ad4d93879308126554df2e565015eb799f62fbfbfcfba5e5b69237fb55a78509","first_computed_at":"2026-07-05T06:18:24.409468Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:18:24.409468Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kt3K7axsY85TNnwdy+uqZh8x22OjmtI6PQZ13dRW/qKYocqvWTMTitWFMIdiUNACSvjg2zROFXE8dddOXRZLCg==","signature_status":"signed_v1","signed_at":"2026-07-05T06:18:24.409923Z","signed_message":"canonical_sha256_bytes"},"source_id":"2306.03968","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d01d93e1e4fd019905f20349e9cd0562620e3790edf9ce29eee85ab01914883b","sha256:e3bb103edbcc715a987ffd67576c8c1b3e47c5ddfd4bc83de0b73205c036ca6f"],"state_sha256":"0b9f5f887c4a9630dbe04b1c09c915da38f0436f697bbb56bbad1ec029e9cf72"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aZ14nFBwzrBVozXXPZtuwJqVtru8T/WyMmWW0txpjpFl/t7cmi8b9hnNs0Juh8BX/8K+FI8QydBaNTffFI0PBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-25T23:09:02.157290Z","bundle_sha256":"0f86e427d0c8d657e8720c72897a9c94957ea979e6666fc6cb2cb192e645e54a"}}