{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:5P3DZCLVV5F5JYST4UFW7GN6AH","short_pith_number":"pith:5P3DZCLV","canonical_record":{"source":{"id":"2310.18913","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T05:50:03Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"23048f0b14c3d5d0470e4a56d28ac44c2942ac1e39c51917925378ded1f7d202","abstract_canon_sha256":"3df0f840be371f2441139710ae91336f81fde9a2322ca778d77d3b57b2f74816"},"schema_version":"1.0"},"canonical_sha256":"ebf63c8975af4bd4e253e50b6f99be01c107521f4ffecd3b54a62fd6ccaa33fe","source":{"kind":"arxiv","id":"2310.18913","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.18913","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"arxiv_version","alias_value":"2310.18913v4","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.18913","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_12","alias_value":"5P3DZCLVV5F5","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_16","alias_value":"5P3DZCLVV5F5JYST","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_8","alias_value":"5P3DZCLV","created_at":"2026-07-05T08:24:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:5P3DZCLVV5F5JYST4UFW7GN6AH","target":"record","payload":{"canonical_record":{"source":{"id":"2310.18913","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T05:50:03Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"23048f0b14c3d5d0470e4a56d28ac44c2942ac1e39c51917925378ded1f7d202","abstract_canon_sha256":"3df0f840be371f2441139710ae91336f81fde9a2322ca778d77d3b57b2f74816"},"schema_version":"1.0"},"canonical_sha256":"ebf63c8975af4bd4e253e50b6f99be01c107521f4ffecd3b54a62fd6ccaa33fe","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:38.186827Z","signature_b64":"Fs7PcAQyxw7cP78uRSIj/6z5DXFTxm2fE4mhP7dvt/8bRjO5qPuMn703HjVJDWnDBbrSAdO8VNrUabQhIH3FAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ebf63c8975af4bd4e253e50b6f99be01c107521f4ffecd3b54a62fd6ccaa33fe","last_reissued_at":"2026-07-05T08:24:38.186303Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:38.186303Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.18913","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:24:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oGUQn6SqSyzpvqdQWCZAyoGdliaZQBpOuegZeUQ9XlCgbxvZ4QZCEBOCtJ70Nat5VP15pB2TMrZ8iI31Jnk8Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:37:28.911227Z"},"content_sha256":"33cb9742c61f4337e1e00299ac020a1f31e32e2a3cdbaaba0a02222358bd82c3","schema_version":"1.0","event_id":"sha256:33cb9742c61f4337e1e00299ac020a1f31e32e2a3cdbaaba0a02222358bd82c3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:5P3DZCLVV5F5JYST4UFW7GN6AH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Debiasing Algorithm through Model Adaptation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.CL","authors_text":"David Mare\\v{c}ek, Tomasz Limisiewicz, Tom\\'a\\v{s} Musil","submitted_at":"2023-10-29T05:50:03Z","abstract_excerpt":"Large language models are becoming the go-to solution for the ever-growing number of tasks. However, with growing capacity, models are prone to rely on spurious correlations stemming from biases and stereotypes present in the training data. This work proposes a novel method for detecting and mitigating gender bias in language models. We perform causal analysis to identify problematic model components and discover that mid-upper feed-forward layers are most prone to convey bias. Based on the analysis results, we intervene in the model by applying a linear projection to the weight matrices of th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.18913","kind":"arxiv","version":4},"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/2310.18913/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":"2026-07-05T08:24:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ASUuhGowCfNlGzTYHjXYXLrod0Gz7SepFJlrSDYFBUy7pdVhfOmBsGJqwhkJeQ60B9N5WnWvC7ojLTr0r9nvBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:37:28.911732Z"},"content_sha256":"cf0071304b67b5bb9a3ff5cf1b204394392e62561fff3a33001ee6e859d028f9","schema_version":"1.0","event_id":"sha256:cf0071304b67b5bb9a3ff5cf1b204394392e62561fff3a33001ee6e859d028f9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/bundle.json","state_url":"https://pith.science/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/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":"2026-08-08T23:37:28Z","links":{"resolver":"https://pith.science/pith/5P3DZCLVV5F5JYST4UFW7GN6AH","bundle":"https://pith.science/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/bundle.json","state":"https://pith.science/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5P3DZCLVV5F5JYST4UFW7GN6AH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:5P3DZCLVV5F5JYST4UFW7GN6AH","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":"3df0f840be371f2441139710ae91336f81fde9a2322ca778d77d3b57b2f74816","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T05:50:03Z","title_canon_sha256":"23048f0b14c3d5d0470e4a56d28ac44c2942ac1e39c51917925378ded1f7d202"},"schema_version":"1.0","source":{"id":"2310.18913","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.18913","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"arxiv_version","alias_value":"2310.18913v4","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.18913","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_12","alias_value":"5P3DZCLVV5F5","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_16","alias_value":"5P3DZCLVV5F5JYST","created_at":"2026-07-05T08:24:38Z"},{"alias_kind":"pith_short_8","alias_value":"5P3DZCLV","created_at":"2026-07-05T08:24:38Z"}],"graph_snapshots":[{"event_id":"sha256:cf0071304b67b5bb9a3ff5cf1b204394392e62561fff3a33001ee6e859d028f9","target":"graph","created_at":"2026-07-05T08:24:38Z","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/2310.18913/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models are becoming the go-to solution for the ever-growing number of tasks. However, with growing capacity, models are prone to rely on spurious correlations stemming from biases and stereotypes present in the training data. This work proposes a novel method for detecting and mitigating gender bias in language models. We perform causal analysis to identify problematic model components and discover that mid-upper feed-forward layers are most prone to convey bias. Based on the analysis results, we intervene in the model by applying a linear projection to the weight matrices of th","authors_text":"David Mare\\v{c}ek, Tomasz Limisiewicz, Tom\\'a\\v{s} Musil","cross_cats":["cs.AI","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T05:50:03Z","title":"Debiasing Algorithm through Model Adaptation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.18913","kind":"arxiv","version":4},"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":"sha256:33cb9742c61f4337e1e00299ac020a1f31e32e2a3cdbaaba0a02222358bd82c3","target":"record","created_at":"2026-07-05T08:24:38Z","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":"3df0f840be371f2441139710ae91336f81fde9a2322ca778d77d3b57b2f74816","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T05:50:03Z","title_canon_sha256":"23048f0b14c3d5d0470e4a56d28ac44c2942ac1e39c51917925378ded1f7d202"},"schema_version":"1.0","source":{"id":"2310.18913","kind":"arxiv","version":4}},"canonical_sha256":"ebf63c8975af4bd4e253e50b6f99be01c107521f4ffecd3b54a62fd6ccaa33fe","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ebf63c8975af4bd4e253e50b6f99be01c107521f4ffecd3b54a62fd6ccaa33fe","first_computed_at":"2026-07-05T08:24:38.186303Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:24:38.186303Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Fs7PcAQyxw7cP78uRSIj/6z5DXFTxm2fE4mhP7dvt/8bRjO5qPuMn703HjVJDWnDBbrSAdO8VNrUabQhIH3FAw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:24:38.186827Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.18913","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:33cb9742c61f4337e1e00299ac020a1f31e32e2a3cdbaaba0a02222358bd82c3","sha256:cf0071304b67b5bb9a3ff5cf1b204394392e62561fff3a33001ee6e859d028f9"],"state_sha256":"64e6d3bb2f4a2be9b27a41784391c4736e65cb29a5e2c87b77331c9a01d7fbbb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"izt337/R9cFJYOHAj8wrnXVDjlEhjcyYwBUcrw1ClJs+m641RO6Yz4AFqe1SD8QhCt26DemjSEpdeujn1NJ7Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T23:37:28.916659Z","bundle_sha256":"43234d0f6de9c7deddbd78325045accd725d486c9d094daf1fc05b95d7a35c01"}}