{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:INC3MH7AWLWQTGSSNYHKL3LYFW","short_pith_number":"pith:INC3MH7A","canonical_record":{"source":{"id":"2110.04020","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T10:35:27Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"8ebe3ec22a2ddfc4f95a971015c9981008c3969ab69781abf1c88962d4883e59","abstract_canon_sha256":"4d4be3fea780fefb5cd1cd47a451e8772d8dd0c11e7e562f252dc49b6ce2b46b"},"schema_version":"1.0"},"canonical_sha256":"4345b61fe0b2ed099a526e0ea5ed782dbc8273398f20ce22d364aeb7d0fc00a6","source":{"kind":"arxiv","id":"2110.04020","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.04020","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"arxiv_version","alias_value":"2110.04020v2","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.04020","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_12","alias_value":"INC3MH7AWLWQ","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_16","alias_value":"INC3MH7AWLWQTGSS","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_8","alias_value":"INC3MH7A","created_at":"2026-07-05T03:22:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:INC3MH7AWLWQTGSSNYHKL3LYFW","target":"record","payload":{"canonical_record":{"source":{"id":"2110.04020","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T10:35:27Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"8ebe3ec22a2ddfc4f95a971015c9981008c3969ab69781abf1c88962d4883e59","abstract_canon_sha256":"4d4be3fea780fefb5cd1cd47a451e8772d8dd0c11e7e562f252dc49b6ce2b46b"},"schema_version":"1.0"},"canonical_sha256":"4345b61fe0b2ed099a526e0ea5ed782dbc8273398f20ce22d364aeb7d0fc00a6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:22:47.339892Z","signature_b64":"kOZeFo55N99jha2P8Sw0Rt5QPX4RHoOpDTGP8vID0fVEwYy2ESy6EguxBeHmp7xGe7Cz9mhugtmaObcf56IqDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4345b61fe0b2ed099a526e0ea5ed782dbc8273398f20ce22d364aeb7d0fc00a6","last_reissued_at":"2026-07-05T03:22:47.339320Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:22:47.339320Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2110.04020","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:22:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"76V3nUhAhaKiYmc1PoBAxsNJOpFOt/DqFkGxKt4jHSmQ2CtPLEmLZtRfXCFVynKVwhg96lx7RPOhdU5DQ6JhBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T16:04:40.277146Z"},"content_sha256":"4ecee58678ef12fc2d3e74e50f587a810e9f67a3c59bdab1ed2590a811f2e4f1","schema_version":"1.0","event_id":"sha256:4ecee58678ef12fc2d3e74e50f587a810e9f67a3c59bdab1ed2590a811f2e4f1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:INC3MH7AWLWQTGSSNYHKL3LYFW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Pathologies in priors and inference for Bayesian transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Alexander Immer, Max Horn, Tristan Cinquin, Vincent Fortuin","submitted_at":"2021-10-08T10:35:27Z","abstract_excerpt":"In recent years, the transformer has established itself as a workhorse in many applications ranging from natural language processing to reinforcement learning. Similarly, Bayesian deep learning has become the gold-standard for uncertainty estimation in safety-critical applications, where robustness and calibration are crucial. Surprisingly, no successful attempts to improve transformer models in terms of predictive uncertainty using Bayesian inference exist. In this work, we study this curiously underpopulated area of Bayesian transformers. We find that weight-space inference in transformers d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.04020","kind":"arxiv","version":2},"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/2110.04020/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":"2026-07-05T03:22:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IXW5ZOUPCUtBr9v0qUDYp754kT2Knlk1v/NJa1PCPyoffC57hYQ5Vabtvuh9nVdvJb7Y5NTHPUY5lliF20XSDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T16:04:40.277673Z"},"content_sha256":"23f23a1e9f151ff717850145935f0d2ab1f010ce9eff9f21e9fc41b6d5b86d66","schema_version":"1.0","event_id":"sha256:23f23a1e9f151ff717850145935f0d2ab1f010ce9eff9f21e9fc41b6d5b86d66"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/bundle.json","state_url":"https://pith.science/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/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":"2026-08-20T16:04:40Z","links":{"resolver":"https://pith.science/pith/INC3MH7AWLWQTGSSNYHKL3LYFW","bundle":"https://pith.science/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/bundle.json","state":"https://pith.science/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/INC3MH7AWLWQTGSSNYHKL3LYFW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:INC3MH7AWLWQTGSSNYHKL3LYFW","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":"4d4be3fea780fefb5cd1cd47a451e8772d8dd0c11e7e562f252dc49b6ce2b46b","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T10:35:27Z","title_canon_sha256":"8ebe3ec22a2ddfc4f95a971015c9981008c3969ab69781abf1c88962d4883e59"},"schema_version":"1.0","source":{"id":"2110.04020","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.04020","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"arxiv_version","alias_value":"2110.04020v2","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.04020","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_12","alias_value":"INC3MH7AWLWQ","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_16","alias_value":"INC3MH7AWLWQTGSS","created_at":"2026-07-05T03:22:47Z"},{"alias_kind":"pith_short_8","alias_value":"INC3MH7A","created_at":"2026-07-05T03:22:47Z"}],"graph_snapshots":[{"event_id":"sha256:23f23a1e9f151ff717850145935f0d2ab1f010ce9eff9f21e9fc41b6d5b86d66","target":"graph","created_at":"2026-07-05T03:22:47Z","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/2110.04020/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In recent years, the transformer has established itself as a workhorse in many applications ranging from natural language processing to reinforcement learning. Similarly, Bayesian deep learning has become the gold-standard for uncertainty estimation in safety-critical applications, where robustness and calibration are crucial. Surprisingly, no successful attempts to improve transformer models in terms of predictive uncertainty using Bayesian inference exist. In this work, we study this curiously underpopulated area of Bayesian transformers. We find that weight-space inference in transformers d","authors_text":"Alexander Immer, Max Horn, Tristan Cinquin, Vincent Fortuin","cross_cats":["stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T10:35:27Z","title":"Pathologies in priors and inference for Bayesian transformers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.04020","kind":"arxiv","version":2},"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":"sha256:4ecee58678ef12fc2d3e74e50f587a810e9f67a3c59bdab1ed2590a811f2e4f1","target":"record","created_at":"2026-07-05T03:22:47Z","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":"4d4be3fea780fefb5cd1cd47a451e8772d8dd0c11e7e562f252dc49b6ce2b46b","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T10:35:27Z","title_canon_sha256":"8ebe3ec22a2ddfc4f95a971015c9981008c3969ab69781abf1c88962d4883e59"},"schema_version":"1.0","source":{"id":"2110.04020","kind":"arxiv","version":2}},"canonical_sha256":"4345b61fe0b2ed099a526e0ea5ed782dbc8273398f20ce22d364aeb7d0fc00a6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4345b61fe0b2ed099a526e0ea5ed782dbc8273398f20ce22d364aeb7d0fc00a6","first_computed_at":"2026-07-05T03:22:47.339320Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:22:47.339320Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"kOZeFo55N99jha2P8Sw0Rt5QPX4RHoOpDTGP8vID0fVEwYy2ESy6EguxBeHmp7xGe7Cz9mhugtmaObcf56IqDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T03:22:47.339892Z","signed_message":"canonical_sha256_bytes"},"source_id":"2110.04020","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4ecee58678ef12fc2d3e74e50f587a810e9f67a3c59bdab1ed2590a811f2e4f1","sha256:23f23a1e9f151ff717850145935f0d2ab1f010ce9eff9f21e9fc41b6d5b86d66"],"state_sha256":"ec660d69b43f5313eea308eb20d5af8d8f7a5e7e7e4adc66857a68eeb5190512"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7dvr1TMcmCbdi1jJry/nrBXhb1KODhlj9Av/EvJbHt/dykwN0fQ3efdTZfEwdE09O3vT6duO5RKIBCPoiW0IAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T16:04:40.281672Z","bundle_sha256":"533fe9ce258ca11ff15df9efefa0543a2cd7e70cce76b727c6de8173c93ba8a4"}}