{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:AYBJSQCIGBSWWH3KHKO3Y2KW7Z","short_pith_number":"pith:AYBJSQCI","canonical_record":{"source":{"id":"2007.04206","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-08T15:48:12Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"f62f03ac61d58e692e48f464739f65885cd7263338b112a8d23ec32ee46d4351","abstract_canon_sha256":"ddeb4b420315e915431eb79996b0038b7d6feadf940fa37857cc46090f6a7b8b"},"schema_version":"1.0"},"canonical_sha256":"060299404830656b1f6a3a9dbc6956fe5019b9ef15a18ab853da9eecd964726a","source":{"kind":"arxiv","id":"2007.04206","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2007.04206","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"arxiv_version","alias_value":"2007.04206v1","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.04206","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_12","alias_value":"AYBJSQCIGBSW","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_16","alias_value":"AYBJSQCIGBSWWH3K","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_8","alias_value":"AYBJSQCI","created_at":"2026-07-05T01:17:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:AYBJSQCIGBSWWH3KHKO3Y2KW7Z","target":"record","payload":{"canonical_record":{"source":{"id":"2007.04206","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-08T15:48:12Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"f62f03ac61d58e692e48f464739f65885cd7263338b112a8d23ec32ee46d4351","abstract_canon_sha256":"ddeb4b420315e915431eb79996b0038b7d6feadf940fa37857cc46090f6a7b8b"},"schema_version":"1.0"},"canonical_sha256":"060299404830656b1f6a3a9dbc6956fe5019b9ef15a18ab853da9eecd964726a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:17:21.999717Z","signature_b64":"krDTNast+UKJsHrAZgAA30ViR2t3MxnSoiqlDy36WTHEudH1tV2CPY7K/ChrGQ/RVEp2daXmliXBg7eXcwGTBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"060299404830656b1f6a3a9dbc6956fe5019b9ef15a18ab853da9eecd964726a","last_reissued_at":"2026-07-05T01:17:21.999309Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:17:21.999309Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2007.04206","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":"2026-07-05T01:17:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OwzJCayU2vCtSj+CxZjfS3+QC2rXdb6nfFXMQOpoCFv7IIc4jX721vNDB4erBLc3HOzbfrhXzd1t4iGfFU/qBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T03:09:07.420277Z"},"content_sha256":"dd0e3b7a0177151795f19f78a4443a8d1c6b9deee217920a87dc4381e043e0c6","schema_version":"1.0","event_id":"sha256:dd0e3b7a0177151795f19f78a4443a8d1c6b9deee217920a87dc4381e043e0c6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:AYBJSQCIGBSWWH3KHKO3Y2KW7Z","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Diverse Ensembles Improve Calibration","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Asa Cooper Stickland, Iain Murray","submitted_at":"2020-07-08T15:48:12Z","abstract_excerpt":"Modern deep neural networks can produce badly calibrated predictions, especially when train and test distributions are mismatched. Training an ensemble of models and averaging their predictions can help alleviate these issues. We propose a simple technique to improve calibration, using a different data augmentation for each ensemble member. We additionally use the idea of `mixing' un-augmented and augmented inputs to improve calibration when test and training distributions are the same. These simple techniques improve calibration and accuracy over strong baselines on the CIFAR10 and CIFAR100 b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.04206","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/2007.04206/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":"2026-07-05T01:17:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FEnsmFtgBbRj7ynlcuPilEkQwLC9CfAMyBqWHw5QgC7Fb6smFth/VjhdXwDxBVirFcxtQ+4HgnByH79muvlWCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T03:09:07.420846Z"},"content_sha256":"d53b273949a89c8d114a5edf02e03a552b3c6f63a531d226abc3fbdf35accffe","schema_version":"1.0","event_id":"sha256:d53b273949a89c8d114a5edf02e03a552b3c6f63a531d226abc3fbdf35accffe"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/bundle.json","state_url":"https://pith.science/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/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":"2026-08-05T03:09:07Z","links":{"resolver":"https://pith.science/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z","bundle":"https://pith.science/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/bundle.json","state":"https://pith.science/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/state.json","well_known_bundle":"https://pith.science/.well-known/pith/AYBJSQCIGBSWWH3KHKO3Y2KW7Z/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:AYBJSQCIGBSWWH3KHKO3Y2KW7Z","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":"ddeb4b420315e915431eb79996b0038b7d6feadf940fa37857cc46090f6a7b8b","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-08T15:48:12Z","title_canon_sha256":"f62f03ac61d58e692e48f464739f65885cd7263338b112a8d23ec32ee46d4351"},"schema_version":"1.0","source":{"id":"2007.04206","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2007.04206","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"arxiv_version","alias_value":"2007.04206v1","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2007.04206","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_12","alias_value":"AYBJSQCIGBSW","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_16","alias_value":"AYBJSQCIGBSWWH3K","created_at":"2026-07-05T01:17:21Z"},{"alias_kind":"pith_short_8","alias_value":"AYBJSQCI","created_at":"2026-07-05T01:17:21Z"}],"graph_snapshots":[{"event_id":"sha256:d53b273949a89c8d114a5edf02e03a552b3c6f63a531d226abc3fbdf35accffe","target":"graph","created_at":"2026-07-05T01:17:21Z","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/2007.04206/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Modern deep neural networks can produce badly calibrated predictions, especially when train and test distributions are mismatched. Training an ensemble of models and averaging their predictions can help alleviate these issues. We propose a simple technique to improve calibration, using a different data augmentation for each ensemble member. We additionally use the idea of `mixing' un-augmented and augmented inputs to improve calibration when test and training distributions are the same. These simple techniques improve calibration and accuracy over strong baselines on the CIFAR10 and CIFAR100 b","authors_text":"Asa Cooper Stickland, Iain Murray","cross_cats":["stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-08T15:48:12Z","title":"Diverse Ensembles Improve Calibration"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2007.04206","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:dd0e3b7a0177151795f19f78a4443a8d1c6b9deee217920a87dc4381e043e0c6","target":"record","created_at":"2026-07-05T01:17:21Z","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":"ddeb4b420315e915431eb79996b0038b7d6feadf940fa37857cc46090f6a7b8b","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2020-07-08T15:48:12Z","title_canon_sha256":"f62f03ac61d58e692e48f464739f65885cd7263338b112a8d23ec32ee46d4351"},"schema_version":"1.0","source":{"id":"2007.04206","kind":"arxiv","version":1}},"canonical_sha256":"060299404830656b1f6a3a9dbc6956fe5019b9ef15a18ab853da9eecd964726a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"060299404830656b1f6a3a9dbc6956fe5019b9ef15a18ab853da9eecd964726a","first_computed_at":"2026-07-05T01:17:21.999309Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T01:17:21.999309Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"krDTNast+UKJsHrAZgAA30ViR2t3MxnSoiqlDy36WTHEudH1tV2CPY7K/ChrGQ/RVEp2daXmliXBg7eXcwGTBw==","signature_status":"signed_v1","signed_at":"2026-07-05T01:17:21.999717Z","signed_message":"canonical_sha256_bytes"},"source_id":"2007.04206","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dd0e3b7a0177151795f19f78a4443a8d1c6b9deee217920a87dc4381e043e0c6","sha256:d53b273949a89c8d114a5edf02e03a552b3c6f63a531d226abc3fbdf35accffe"],"state_sha256":"90ef2eb5cdfd9581b20eaf272748fb4c2e6f3863bd81a9d74dcdb6d58a886704"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+C7hqcaiKn8sclwZPXrVqgvR+QlV94ChTgLaIbaZycSieKgHeASBIAJIzB4LACqCmCS0t6e06nHGK4PHA3qDCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T03:09:07.426303Z","bundle_sha256":"fcd056da0f2253849fc66204be5863c2a07a6abd1f6879420ab750246bae16b9"}}