{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:2W2PKDSRHH6F6NNJEUM6F7XC7R","short_pith_number":"pith:2W2PKDSR","canonical_record":{"source":{"id":"2310.12611","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-19T09:39:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6fd85ee631e6a93b0119e4eeda71eec088b676b3a596453b276c7fd0be766319","abstract_canon_sha256":"090e1035b1424b6cfc8b594043cb6a9d377c62358a7a7aa50cb63ed8534935eb"},"schema_version":"1.0"},"canonical_sha256":"d5b4f50e5139fc5f35a92519e2fee2fc7ce3e651e1d508217525e2510f6e719b","source":{"kind":"arxiv","id":"2310.12611","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.12611","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"arxiv_version","alias_value":"2310.12611v1","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.12611","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_12","alias_value":"2W2PKDSRHH6F","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_16","alias_value":"2W2PKDSRHH6F6NNJ","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_8","alias_value":"2W2PKDSR","created_at":"2026-07-05T07:02:40Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:2W2PKDSRHH6F6NNJEUM6F7XC7R","target":"record","payload":{"canonical_record":{"source":{"id":"2310.12611","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-19T09:39:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6fd85ee631e6a93b0119e4eeda71eec088b676b3a596453b276c7fd0be766319","abstract_canon_sha256":"090e1035b1424b6cfc8b594043cb6a9d377c62358a7a7aa50cb63ed8534935eb"},"schema_version":"1.0"},"canonical_sha256":"d5b4f50e5139fc5f35a92519e2fee2fc7ce3e651e1d508217525e2510f6e719b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:02:40.576394Z","signature_b64":"ZZFhAkiGaIJhhF7AM1suZOdKNoeqClSYRMC8gixs52ksSXgL4slbZOqARDgtbI+ilB+DbuMak/ykKrkdKw/DAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d5b4f50e5139fc5f35a92519e2fee2fc7ce3e651e1d508217525e2510f6e719b","last_reissued_at":"2026-07-05T07:02:40.575981Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:02:40.575981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.12611","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":"2026-07-05T07:02:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VhqCoMDaiUGx4fYX275YcT0fbAPA4oBj1wvW6hk6uVd+s/niShB7vFbjHUz4mFPsp1lWbFDXgAk4nzuAgLXZCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:40:11.371477Z"},"content_sha256":"0d9106caa5f2dee5ea8cd5f3b991f930f4a793802c709e6d0748b67172f7204f","schema_version":"1.0","event_id":"sha256:0d9106caa5f2dee5ea8cd5f3b991f930f4a793802c709e6d0748b67172f7204f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:2W2PKDSRHH6F6NNJEUM6F7XC7R","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Identifying and Adapting Transformer-Components Responsible for Gender Bias in an English Language Model","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abhijith Chintam, Michael Hanna, Oskar van der Wal, Rahel Beloch, Willem Zuidema","submitted_at":"2023-10-19T09:39:21Z","abstract_excerpt":"Language models (LMs) exhibit and amplify many types of undesirable biases learned from the training data, including gender bias. However, we lack tools for effectively and efficiently changing this behavior without hurting general language modeling performance. In this paper, we study three methods for identifying causal relations between LM components and particular output: causal mediation analysis, automated circuit discovery and our novel, efficient method called DiffMask+ based on differential masking. We apply the methods to GPT-2 small and the problem of gender bias, and use the discov"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.12611","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/2310.12611/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":"2026-07-05T07:02:40Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P+YwUoeFhsLh8hNHiki7T/6irt8dlZgXSUtKwRIiPGGhpdgSWYvK8VtR1H5GetW6PtaBZ5d/jd7iloTtKhdOAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T02:40:11.371976Z"},"content_sha256":"2ab48c10e0ff08952e739386d1c80b534b33fd6494d55e3edf727500a9ffcffc","schema_version":"1.0","event_id":"sha256:2ab48c10e0ff08952e739386d1c80b534b33fd6494d55e3edf727500a9ffcffc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/bundle.json","state_url":"https://pith.science/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/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":"2026-08-09T02:40:11Z","links":{"resolver":"https://pith.science/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R","bundle":"https://pith.science/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/bundle.json","state":"https://pith.science/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/state.json","well_known_bundle":"https://pith.science/.well-known/pith/2W2PKDSRHH6F6NNJEUM6F7XC7R/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:2W2PKDSRHH6F6NNJEUM6F7XC7R","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":"090e1035b1424b6cfc8b594043cb6a9d377c62358a7a7aa50cb63ed8534935eb","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-19T09:39:21Z","title_canon_sha256":"6fd85ee631e6a93b0119e4eeda71eec088b676b3a596453b276c7fd0be766319"},"schema_version":"1.0","source":{"id":"2310.12611","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.12611","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"arxiv_version","alias_value":"2310.12611v1","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.12611","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_12","alias_value":"2W2PKDSRHH6F","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_16","alias_value":"2W2PKDSRHH6F6NNJ","created_at":"2026-07-05T07:02:40Z"},{"alias_kind":"pith_short_8","alias_value":"2W2PKDSR","created_at":"2026-07-05T07:02:40Z"}],"graph_snapshots":[{"event_id":"sha256:2ab48c10e0ff08952e739386d1c80b534b33fd6494d55e3edf727500a9ffcffc","target":"graph","created_at":"2026-07-05T07:02:40Z","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/2310.12611/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language models (LMs) exhibit and amplify many types of undesirable biases learned from the training data, including gender bias. However, we lack tools for effectively and efficiently changing this behavior without hurting general language modeling performance. In this paper, we study three methods for identifying causal relations between LM components and particular output: causal mediation analysis, automated circuit discovery and our novel, efficient method called DiffMask+ based on differential masking. We apply the methods to GPT-2 small and the problem of gender bias, and use the discov","authors_text":"Abhijith Chintam, Michael Hanna, Oskar van der Wal, Rahel Beloch, Willem Zuidema","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-19T09:39:21Z","title":"Identifying and Adapting Transformer-Components Responsible for Gender Bias in an English Language Model"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.12611","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":"sha256:0d9106caa5f2dee5ea8cd5f3b991f930f4a793802c709e6d0748b67172f7204f","target":"record","created_at":"2026-07-05T07:02:40Z","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":"090e1035b1424b6cfc8b594043cb6a9d377c62358a7a7aa50cb63ed8534935eb","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-19T09:39:21Z","title_canon_sha256":"6fd85ee631e6a93b0119e4eeda71eec088b676b3a596453b276c7fd0be766319"},"schema_version":"1.0","source":{"id":"2310.12611","kind":"arxiv","version":1}},"canonical_sha256":"d5b4f50e5139fc5f35a92519e2fee2fc7ce3e651e1d508217525e2510f6e719b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d5b4f50e5139fc5f35a92519e2fee2fc7ce3e651e1d508217525e2510f6e719b","first_computed_at":"2026-07-05T07:02:40.575981Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:02:40.575981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZZFhAkiGaIJhhF7AM1suZOdKNoeqClSYRMC8gixs52ksSXgL4slbZOqARDgtbI+ilB+DbuMak/ykKrkdKw/DAg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:02:40.576394Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.12611","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0d9106caa5f2dee5ea8cd5f3b991f930f4a793802c709e6d0748b67172f7204f","sha256:2ab48c10e0ff08952e739386d1c80b534b33fd6494d55e3edf727500a9ffcffc"],"state_sha256":"882caa1bd7716484e20af54080f6d0c8724d48f676bb2770b00e527e36f85b11"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"a3Kb1BSEKWIeukjwk9S/p53BeUi14CG4r0pubM7B5m81PjaXYC4XSByDnVfGlbbhZdAF+n4AypTYuem69O75Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T02:40:11.375218Z","bundle_sha256":"0dee06c6d68193c2a9a03c6bdd5dc69da1b6e54fbc34c4ae0516f9459ca500f1"}}