{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:MRSVEPZX2NFPNDTA4GCUFZRLSA","short_pith_number":"pith:MRSVEPZX","canonical_record":{"source":{"id":"2104.04975","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-04-11T09:50:24Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2a71b16c3df8c072b7d27eb2edf462c17f46fc85161cb54171811967fe81c70f","abstract_canon_sha256":"ca94bb9ef3da582863f8b4cdec6773e4babe657c0ef00045b92ad48525bfc308"},"schema_version":"1.0"},"canonical_sha256":"6465523f37d34af68e60e18542e62b902afc5f1cc8be54fbdcb24cdfcec54faa","source":{"kind":"arxiv","id":"2104.04975","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2104.04975","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"arxiv_version","alias_value":"2104.04975v3","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.04975","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_12","alias_value":"MRSVEPZX2NFP","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_16","alias_value":"MRSVEPZX2NFPNDTA","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_8","alias_value":"MRSVEPZX","created_at":"2026-07-05T02:49:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:MRSVEPZX2NFPNDTA4GCUFZRLSA","target":"record","payload":{"canonical_record":{"source":{"id":"2104.04975","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-04-11T09:50:24Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"2a71b16c3df8c072b7d27eb2edf462c17f46fc85161cb54171811967fe81c70f","abstract_canon_sha256":"ca94bb9ef3da582863f8b4cdec6773e4babe657c0ef00045b92ad48525bfc308"},"schema_version":"1.0"},"canonical_sha256":"6465523f37d34af68e60e18542e62b902afc5f1cc8be54fbdcb24cdfcec54faa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:49:29.388109Z","signature_b64":"spT1Uem7dVG6Ydy1JF2cmYk58Oje7ETqRJU7+jUEYy6ULKHWwGbwWUZywDLXQKGXf6ddAkiZFvVXVlsP1AXaDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6465523f37d34af68e60e18542e62b902afc5f1cc8be54fbdcb24cdfcec54faa","last_reissued_at":"2026-07-05T02:49:29.387641Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:49:29.387641Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2104.04975","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":"2026-07-05T02:49:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hfhFGaIllxB3972WbjrqL94sJn8LlWbVF1MlhnaCP+PjBFrx6NnhiLWDhini7RBfZPaanDOyNvWCa9yxiCJ4Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:31:53.436018Z"},"content_sha256":"05c8b6778a8883e3f060a31494c780898502d81584de86e42e138c67b39a555a","schema_version":"1.0","event_id":"sha256:05c8b6778a8883e3f060a31494c780898502d81584de86e42e138c67b39a555a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:MRSVEPZX2NFPNDTA4GCUFZRLSA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Alexander Immer, Gunnar R\\\"atsch, Matthias Bauer, Mohammad Emtiyaz Khan, Vincent Fortuin","submitted_at":"2021-04-11T09:50:24Z","abstract_excerpt":"Marginal-likelihood based model-selection, even though promising, is rarely used in deep learning due to estimation difficulties. Instead, most approaches rely on validation data, which may not be readily available. In this work, we present a scalable marginal-likelihood estimation method to select both hyperparameters and network architectures, based on the training data alone. Some hyperparameters can be estimated online during training, simplifying the procedure. Our marginal-likelihood estimate is based on Laplace's method and Gauss-Newton approximations to the Hessian, and it outperforms "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.04975","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2104.04975/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":"2026-07-05T02:49:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PQJ5F9mvoTH8nuWUhGOxd3lAue44enfKyBwiIY8hPGs1WQEdbpV54lIsr9eIavJS8KnKz0O9CJcvxrBRq7shDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T02:31:53.436687Z"},"content_sha256":"78104334bcde836d07fa9377e1f7dce67f44306b60c157080a268f480c1d67c4","schema_version":"1.0","event_id":"sha256:78104334bcde836d07fa9377e1f7dce67f44306b60c157080a268f480c1d67c4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/bundle.json","state_url":"https://pith.science/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/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":"2026-08-05T02:31:53Z","links":{"resolver":"https://pith.science/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA","bundle":"https://pith.science/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/bundle.json","state":"https://pith.science/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MRSVEPZX2NFPNDTA4GCUFZRLSA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:MRSVEPZX2NFPNDTA4GCUFZRLSA","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":"ca94bb9ef3da582863f8b4cdec6773e4babe657c0ef00045b92ad48525bfc308","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-04-11T09:50:24Z","title_canon_sha256":"2a71b16c3df8c072b7d27eb2edf462c17f46fc85161cb54171811967fe81c70f"},"schema_version":"1.0","source":{"id":"2104.04975","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2104.04975","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"arxiv_version","alias_value":"2104.04975v3","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.04975","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_12","alias_value":"MRSVEPZX2NFP","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_16","alias_value":"MRSVEPZX2NFPNDTA","created_at":"2026-07-05T02:49:29Z"},{"alias_kind":"pith_short_8","alias_value":"MRSVEPZX","created_at":"2026-07-05T02:49:29Z"}],"graph_snapshots":[{"event_id":"sha256:78104334bcde836d07fa9377e1f7dce67f44306b60c157080a268f480c1d67c4","target":"graph","created_at":"2026-07-05T02:49:29Z","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/2104.04975/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Marginal-likelihood based model-selection, even though promising, is rarely used in deep learning due to estimation difficulties. Instead, most approaches rely on validation data, which may not be readily available. In this work, we present a scalable marginal-likelihood estimation method to select both hyperparameters and network architectures, based on the training data alone. Some hyperparameters can be estimated online during training, simplifying the procedure. Our marginal-likelihood estimate is based on Laplace's method and Gauss-Newton approximations to the Hessian, and it outperforms ","authors_text":"Alexander Immer, Gunnar R\\\"atsch, Matthias Bauer, Mohammad Emtiyaz Khan, Vincent Fortuin","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-04-11T09:50:24Z","title":"Scalable Marginal Likelihood Estimation for Model Selection in Deep Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.04975","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":"sha256:05c8b6778a8883e3f060a31494c780898502d81584de86e42e138c67b39a555a","target":"record","created_at":"2026-07-05T02:49:29Z","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":"ca94bb9ef3da582863f8b4cdec6773e4babe657c0ef00045b92ad48525bfc308","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2021-04-11T09:50:24Z","title_canon_sha256":"2a71b16c3df8c072b7d27eb2edf462c17f46fc85161cb54171811967fe81c70f"},"schema_version":"1.0","source":{"id":"2104.04975","kind":"arxiv","version":3}},"canonical_sha256":"6465523f37d34af68e60e18542e62b902afc5f1cc8be54fbdcb24cdfcec54faa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6465523f37d34af68e60e18542e62b902afc5f1cc8be54fbdcb24cdfcec54faa","first_computed_at":"2026-07-05T02:49:29.387641Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:49:29.387641Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"spT1Uem7dVG6Ydy1JF2cmYk58Oje7ETqRJU7+jUEYy6ULKHWwGbwWUZywDLXQKGXf6ddAkiZFvVXVlsP1AXaDw==","signature_status":"signed_v1","signed_at":"2026-07-05T02:49:29.388109Z","signed_message":"canonical_sha256_bytes"},"source_id":"2104.04975","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:05c8b6778a8883e3f060a31494c780898502d81584de86e42e138c67b39a555a","sha256:78104334bcde836d07fa9377e1f7dce67f44306b60c157080a268f480c1d67c4"],"state_sha256":"87f9aa989246f3b40177a9992072fe6074b7e88c8c9aa3f430d8575b996d329b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2PvMExKWrLEn3yJ0JeMtpeR6NRsH58mYx/bIuhFpW2iw+84pXaxa3Xd8+fjL3Y5KSugeLAlLouVkQ7fV/h9QBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T02:31:53.443278Z","bundle_sha256":"e8cd3e35546c4d858586bd72de636722e828e75d44ce6942cb13825f89c91d67"}}