{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:GVNEY46TQ7YUKMVCINVLZHXGQZ","short_pith_number":"pith:GVNEY46T","canonical_record":{"source":{"id":"2406.16846","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-24T17:51:01Z","cross_cats_sorted":["cs.CY","stat.ML"],"title_canon_sha256":"d63e8147a86bde59b189690500fae78b929f23a344c3eac736519a05599af65a","abstract_canon_sha256":"2c00709f9a3ec1a4700633c3c496baa9969dd9d84d231d801e2b332e972ffc82"},"schema_version":"1.0"},"canonical_sha256":"355a4c73d387f14532a2436abc9ee6867e77bab56a36eef353e0ed13abdae537","source":{"kind":"arxiv","id":"2406.16846","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.16846","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"arxiv_version","alias_value":"2406.16846v1","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.16846","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_12","alias_value":"GVNEY46TQ7YU","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_16","alias_value":"GVNEY46TQ7YUKMVC","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_8","alias_value":"GVNEY46T","created_at":"2026-07-05T08:36:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:GVNEY46TQ7YUKMVCINVLZHXGQZ","target":"record","payload":{"canonical_record":{"source":{"id":"2406.16846","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-24T17:51:01Z","cross_cats_sorted":["cs.CY","stat.ML"],"title_canon_sha256":"d63e8147a86bde59b189690500fae78b929f23a344c3eac736519a05599af65a","abstract_canon_sha256":"2c00709f9a3ec1a4700633c3c496baa9969dd9d84d231d801e2b332e972ffc82"},"schema_version":"1.0"},"canonical_sha256":"355a4c73d387f14532a2436abc9ee6867e77bab56a36eef353e0ed13abdae537","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:04.442661Z","signature_b64":"5Nv7rsrp290WOscOROvpRxH6VotVwcB5dI7GOxv9EofueK7YrQCkmLUD+QIt/5RVGaRZORFilEAltN0E/XSJBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"355a4c73d387f14532a2436abc9ee6867e77bab56a36eef353e0ed13abdae537","last_reissued_at":"2026-07-05T08:36:04.442111Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:04.442111Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.16846","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":"2026-07-05T08:36:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TAoFSY+hpkI0ywbJa9MXhpMEu4MrfzZ9XCIsvQEdqe5JWEt7fYZjjsNRofmhXQK89FRHl29aGAs8vLsajNu8Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T03:01:01.594624Z"},"content_sha256":"c84d5207bc0a6ec05f1e50249f8e389553890c0e9a9dc94bd670532874eda585","schema_version":"1.0","event_id":"sha256:c84d5207bc0a6ec05f1e50249f8e389553890c0e9a9dc94bd670532874eda585"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:GVNEY46TQ7YUKMVCINVLZHXGQZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Data Debiasing with Datamodels (D3M): Improving Subgroup Robustness via Data Selection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CY","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aleksander Madry, Andrew Ilyas, Kimia Hamidieh, Kristian Georgiev, Marzyeh Ghassemi, Saachi Jain","submitted_at":"2024-06-24T17:51:01Z","abstract_excerpt":"Machine learning models can fail on subgroups that are underrepresented during training. While techniques such as dataset balancing can improve performance on underperforming groups, they require access to training group annotations and can end up removing large portions of the dataset. In this paper, we introduce Data Debiasing with Datamodels (D3M), a debiasing approach which isolates and removes specific training examples that drive the model's failures on minority groups. Our approach enables us to efficiently train debiased classifiers while removing only a small number of examples, and d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.16846","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/2406.16846/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":"2026-07-05T08:36:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AryhYe0rkCwMcpM2w8fT2tL1j/OKYiVUcdWABqzSrVc7E9+cRqTzlMK8YT4VobCACtDZb6fGZO2Z6OZWC3iqDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T03:01:01.594992Z"},"content_sha256":"f09f273d1647ceebe67f11a97a2eb5c98b3da975e4a86383a7bd101357e6f9cb","schema_version":"1.0","event_id":"sha256:f09f273d1647ceebe67f11a97a2eb5c98b3da975e4a86383a7bd101357e6f9cb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/bundle.json","state_url":"https://pith.science/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/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":"2026-07-25T03:01:01Z","links":{"resolver":"https://pith.science/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ","bundle":"https://pith.science/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/bundle.json","state":"https://pith.science/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GVNEY46TQ7YUKMVCINVLZHXGQZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:GVNEY46TQ7YUKMVCINVLZHXGQZ","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":"2c00709f9a3ec1a4700633c3c496baa9969dd9d84d231d801e2b332e972ffc82","cross_cats_sorted":["cs.CY","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-24T17:51:01Z","title_canon_sha256":"d63e8147a86bde59b189690500fae78b929f23a344c3eac736519a05599af65a"},"schema_version":"1.0","source":{"id":"2406.16846","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.16846","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"arxiv_version","alias_value":"2406.16846v1","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.16846","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_12","alias_value":"GVNEY46TQ7YU","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_16","alias_value":"GVNEY46TQ7YUKMVC","created_at":"2026-07-05T08:36:04Z"},{"alias_kind":"pith_short_8","alias_value":"GVNEY46T","created_at":"2026-07-05T08:36:04Z"}],"graph_snapshots":[{"event_id":"sha256:f09f273d1647ceebe67f11a97a2eb5c98b3da975e4a86383a7bd101357e6f9cb","target":"graph","created_at":"2026-07-05T08:36:04Z","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/2406.16846/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Machine learning models can fail on subgroups that are underrepresented during training. While techniques such as dataset balancing can improve performance on underperforming groups, they require access to training group annotations and can end up removing large portions of the dataset. In this paper, we introduce Data Debiasing with Datamodels (D3M), a debiasing approach which isolates and removes specific training examples that drive the model's failures on minority groups. Our approach enables us to efficiently train debiased classifiers while removing only a small number of examples, and d","authors_text":"Aleksander Madry, Andrew Ilyas, Kimia Hamidieh, Kristian Georgiev, Marzyeh Ghassemi, Saachi Jain","cross_cats":["cs.CY","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-24T17:51:01Z","title":"Data Debiasing with Datamodels (D3M): Improving Subgroup Robustness via Data Selection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.16846","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:c84d5207bc0a6ec05f1e50249f8e389553890c0e9a9dc94bd670532874eda585","target":"record","created_at":"2026-07-05T08:36:04Z","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":"2c00709f9a3ec1a4700633c3c496baa9969dd9d84d231d801e2b332e972ffc82","cross_cats_sorted":["cs.CY","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-06-24T17:51:01Z","title_canon_sha256":"d63e8147a86bde59b189690500fae78b929f23a344c3eac736519a05599af65a"},"schema_version":"1.0","source":{"id":"2406.16846","kind":"arxiv","version":1}},"canonical_sha256":"355a4c73d387f14532a2436abc9ee6867e77bab56a36eef353e0ed13abdae537","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"355a4c73d387f14532a2436abc9ee6867e77bab56a36eef353e0ed13abdae537","first_computed_at":"2026-07-05T08:36:04.442111Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:36:04.442111Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5Nv7rsrp290WOscOROvpRxH6VotVwcB5dI7GOxv9EofueK7YrQCkmLUD+QIt/5RVGaRZORFilEAltN0E/XSJBA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:36:04.442661Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.16846","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c84d5207bc0a6ec05f1e50249f8e389553890c0e9a9dc94bd670532874eda585","sha256:f09f273d1647ceebe67f11a97a2eb5c98b3da975e4a86383a7bd101357e6f9cb"],"state_sha256":"9d0c4661fe18365b5247417e1de9c02639aad1d0b220a6f6982a90f8af976520"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HBTt5iizKq3MK1F5jdRqAJnt57+WtjtvOOR1mYlCXKOMPyqd8nUWZD9+MIZbgm0emr4mxAGR88c29pMOxJsBDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-25T03:01:01.597431Z","bundle_sha256":"bbcf5cb5eeb4f51ad7bf01e23d19cc86c95ba01f385353fb68abe259eca8cbc5"}}