{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:GRLAU6I5YJTFY2AKYHIOOPTKE4","short_pith_number":"pith:GRLAU6I5","canonical_record":{"source":{"id":"2207.08982","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-07-18T23:43:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a9b2f94d70fca2681a670018b18c264a8e2d43e16a480e103587bed6291105a5","abstract_canon_sha256":"e100fccaf1e44276d133923fac146eee30d3bfc0024867e38920a4a63b464f94"},"schema_version":"1.0"},"canonical_sha256":"34560a791dc2665c680ac1d0e73e6a271e46170a858463bd8ad3711f8ed56249","source":{"kind":"arxiv","id":"2207.08982","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2207.08982","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"arxiv_version","alias_value":"2207.08982v1","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.08982","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_12","alias_value":"GRLAU6I5YJTF","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_16","alias_value":"GRLAU6I5YJTFY2AK","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_8","alias_value":"GRLAU6I5","created_at":"2026-07-05T04:41:15Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:GRLAU6I5YJTFY2AKYHIOOPTKE4","target":"record","payload":{"canonical_record":{"source":{"id":"2207.08982","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-07-18T23:43:52Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a9b2f94d70fca2681a670018b18c264a8e2d43e16a480e103587bed6291105a5","abstract_canon_sha256":"e100fccaf1e44276d133923fac146eee30d3bfc0024867e38920a4a63b464f94"},"schema_version":"1.0"},"canonical_sha256":"34560a791dc2665c680ac1d0e73e6a271e46170a858463bd8ad3711f8ed56249","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:41:15.523407Z","signature_b64":"LhpJ2eKwJKfXb53PXZIlyEAlgyvJ1cBv1KcqMm9sZifDOQyroDlgQw3QJRNGbn8T8khb7jCy+JBsWxdpQ5dLAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34560a791dc2665c680ac1d0e73e6a271e46170a858463bd8ad3711f8ed56249","last_reissued_at":"2026-07-05T04:41:15.522973Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:41:15.522973Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2207.08982","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":"2026-07-05T04:41:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cTKb+XbOARDkX/zHum8x78ZngrNiuKTZcTdnZvmLrYSujmd7Ibe4yp7a5t6vbF3AsVoMSs5EM/3n/ICnrRKLCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T16:52:40.909674Z"},"content_sha256":"39bef4bd655ff1afaf3c7098a0236275165a2e33b4b1937f3bb7e819e1124970","schema_version":"1.0","event_id":"sha256:39bef4bd655ff1afaf3c7098a0236275165a2e33b4b1937f3bb7e819e1124970"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:GRLAU6I5YJTFY2AKYHIOOPTKE4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Selection Bias Induced Spurious Correlations in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Emily McMilin","submitted_at":"2022-07-18T23:43:52Z","abstract_excerpt":"In this work we show how large language models (LLMs) can learn statistical dependencies between otherwise unconditionally independent variables due to dataset selection bias. To demonstrate the effect, we developed a masked gender task that can be applied to BERT-family models to reveal spurious correlations between predicted gender pronouns and a variety of seemingly gender-neutral variables like date and location, on pre-trained (unmodified) BERT and RoBERTa large models. Finally, we provide an online demo, inviting readers to experiment further."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.08982","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/2207.08982/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":"2026-07-05T04:41:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E51Cq5S95s4K1lX23mZT0DIUylhSg+vsJPkaeclwIvjnXz/pKswVm329TjeyVUwgAWomBUNGFMB9qJb9IH4kCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T16:52:40.910204Z"},"content_sha256":"2379cc446cbc7ab02da403f0a0639cc686b0543933053d280642d2872a5eb9dc","schema_version":"1.0","event_id":"sha256:2379cc446cbc7ab02da403f0a0639cc686b0543933053d280642d2872a5eb9dc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/bundle.json","state_url":"https://pith.science/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/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":"2026-08-22T16:52:40Z","links":{"resolver":"https://pith.science/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4","bundle":"https://pith.science/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/bundle.json","state":"https://pith.science/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GRLAU6I5YJTFY2AKYHIOOPTKE4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:GRLAU6I5YJTFY2AKYHIOOPTKE4","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":"e100fccaf1e44276d133923fac146eee30d3bfc0024867e38920a4a63b464f94","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-07-18T23:43:52Z","title_canon_sha256":"a9b2f94d70fca2681a670018b18c264a8e2d43e16a480e103587bed6291105a5"},"schema_version":"1.0","source":{"id":"2207.08982","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2207.08982","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"arxiv_version","alias_value":"2207.08982v1","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.08982","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_12","alias_value":"GRLAU6I5YJTF","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_16","alias_value":"GRLAU6I5YJTFY2AK","created_at":"2026-07-05T04:41:15Z"},{"alias_kind":"pith_short_8","alias_value":"GRLAU6I5","created_at":"2026-07-05T04:41:15Z"}],"graph_snapshots":[{"event_id":"sha256:2379cc446cbc7ab02da403f0a0639cc686b0543933053d280642d2872a5eb9dc","target":"graph","created_at":"2026-07-05T04:41:15Z","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/2207.08982/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In this work we show how large language models (LLMs) can learn statistical dependencies between otherwise unconditionally independent variables due to dataset selection bias. To demonstrate the effect, we developed a masked gender task that can be applied to BERT-family models to reveal spurious correlations between predicted gender pronouns and a variety of seemingly gender-neutral variables like date and location, on pre-trained (unmodified) BERT and RoBERTa large models. Finally, we provide an online demo, inviting readers to experiment further.","authors_text":"Emily McMilin","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-07-18T23:43:52Z","title":"Selection Bias Induced Spurious Correlations in Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.08982","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":"sha256:39bef4bd655ff1afaf3c7098a0236275165a2e33b4b1937f3bb7e819e1124970","target":"record","created_at":"2026-07-05T04:41:15Z","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":"e100fccaf1e44276d133923fac146eee30d3bfc0024867e38920a4a63b464f94","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-07-18T23:43:52Z","title_canon_sha256":"a9b2f94d70fca2681a670018b18c264a8e2d43e16a480e103587bed6291105a5"},"schema_version":"1.0","source":{"id":"2207.08982","kind":"arxiv","version":1}},"canonical_sha256":"34560a791dc2665c680ac1d0e73e6a271e46170a858463bd8ad3711f8ed56249","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"34560a791dc2665c680ac1d0e73e6a271e46170a858463bd8ad3711f8ed56249","first_computed_at":"2026-07-05T04:41:15.522973Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:41:15.522973Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"LhpJ2eKwJKfXb53PXZIlyEAlgyvJ1cBv1KcqMm9sZifDOQyroDlgQw3QJRNGbn8T8khb7jCy+JBsWxdpQ5dLAA==","signature_status":"signed_v1","signed_at":"2026-07-05T04:41:15.523407Z","signed_message":"canonical_sha256_bytes"},"source_id":"2207.08982","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:39bef4bd655ff1afaf3c7098a0236275165a2e33b4b1937f3bb7e819e1124970","sha256:2379cc446cbc7ab02da403f0a0639cc686b0543933053d280642d2872a5eb9dc"],"state_sha256":"827d333020fab2a402ef32a6ccc21f622a2fec0ac5f16c914cec837831cff318"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Yi04/dht55u/H/+tPqduf8Ek700//gCTrMcnHGbLMHLQbiM9CsI/EhWNk7hX04j4sWJWo8SCD3dIysyS+CFfAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-22T16:52:40.914326Z","bundle_sha256":"4cf98516b628df9765810da682a9a0e348fa5e836ce8fa485306a18809366644"}}