{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:RKGRAZWWPKOH6X3GS5VDIGWVUG","short_pith_number":"pith:RKGRAZWW","canonical_record":{"source":{"id":"2412.02893","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-03T22:58:21Z","cross_cats_sorted":["cs.AI","cs.LG","stat.AP","stat.ME"],"title_canon_sha256":"b913bfcfb79b84f676770e49418e153b2af78f1b06c4e854b6291d152cfa8602","abstract_canon_sha256":"a172c9a7789826c11321e47d30d72ba5c0ea8ae9a7333155c85f2c1a6f83c8b8"},"schema_version":"1.0"},"canonical_sha256":"8a8d1066d67a9c7f5f66976a341ad5a1b264e931fb1de4a5b89605127a463da5","source":{"kind":"arxiv","id":"2412.02893","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.02893","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"arxiv_version","alias_value":"2412.02893v1","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.02893","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_12","alias_value":"RKGRAZWWPKOH","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_16","alias_value":"RKGRAZWWPKOH6X3G","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_8","alias_value":"RKGRAZWW","created_at":"2026-07-05T09:44:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:RKGRAZWWPKOH6X3GS5VDIGWVUG","target":"record","payload":{"canonical_record":{"source":{"id":"2412.02893","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-03T22:58:21Z","cross_cats_sorted":["cs.AI","cs.LG","stat.AP","stat.ME"],"title_canon_sha256":"b913bfcfb79b84f676770e49418e153b2af78f1b06c4e854b6291d152cfa8602","abstract_canon_sha256":"a172c9a7789826c11321e47d30d72ba5c0ea8ae9a7333155c85f2c1a6f83c8b8"},"schema_version":"1.0"},"canonical_sha256":"8a8d1066d67a9c7f5f66976a341ad5a1b264e931fb1de4a5b89605127a463da5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:44:21.564546Z","signature_b64":"d+Bs6eKfdTu644azRU/kd7TjxisSkm5y2ug+NQxp8L6E61arCcpvl6gcOAUlYHceU7zsD94gkFLPGJxVRdIxDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a8d1066d67a9c7f5f66976a341ad5a1b264e931fb1de4a5b89605127a463da5","last_reissued_at":"2026-07-05T09:44:21.563983Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:44:21.563983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.02893","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":"2026-07-05T09:44:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UsZSCXP+rxnSRDufkedYOpkJNyyd46ILF78tVhJJ9rHi2HVPBpfiSl+oaOp8sbQ62AW4l/fOQ82IBE904GEQDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T09:33:02.122492Z"},"content_sha256":"0068cf3900b08b5dcc5b4ed1ebcaebff49d6241a401af0ff984483f8d85b8b92","schema_version":"1.0","event_id":"sha256:0068cf3900b08b5dcc5b4ed1ebcaebff49d6241a401af0ff984483f8d85b8b92"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:RKGRAZWWPKOH6X3GS5VDIGWVUG","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Removing Spurious Correlation from Neural Network Interpretations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","stat.AP","stat.ME"],"primary_cat":"cs.CL","authors_text":"David Heckerman, Milad Fotouhi, Mohammad Taha Bahadori, Oluwaseyi Feyisetan, Payman Arabshahi","submitted_at":"2024-12-03T22:58:21Z","abstract_excerpt":"The existing algorithms for identification of neurons responsible for undesired and harmful behaviors do not consider the effects of confounders such as topic of the conversation. In this work, we show that confounders can create spurious correlations and propose a new causal mediation approach that controls the impact of the topic. In experiments with two large language models, we study the localization hypothesis and show that adjusting for the effect of conversation topic, toxicity becomes less localized."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.02893","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/2412.02893/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":"2026-07-05T09:44:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LWXLUW8XKxuvX4JdVyIvHrz0QL3x4YdQ7B9CckZ0Wy6E6+bY+Q9S7u7rGnu2TwztjQRnqU4H/n3GbuMUPk0MBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T09:33:02.123279Z"},"content_sha256":"21619d003a39e6f0cbed14beb5cc62c0c42a5053b5b0a095cc0d375c10e6656e","schema_version":"1.0","event_id":"sha256:21619d003a39e6f0cbed14beb5cc62c0c42a5053b5b0a095cc0d375c10e6656e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/bundle.json","state_url":"https://pith.science/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/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":"2026-08-16T09:33:02Z","links":{"resolver":"https://pith.science/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG","bundle":"https://pith.science/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/bundle.json","state":"https://pith.science/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RKGRAZWWPKOH6X3GS5VDIGWVUG/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:RKGRAZWWPKOH6X3GS5VDIGWVUG","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":"a172c9a7789826c11321e47d30d72ba5c0ea8ae9a7333155c85f2c1a6f83c8b8","cross_cats_sorted":["cs.AI","cs.LG","stat.AP","stat.ME"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-03T22:58:21Z","title_canon_sha256":"b913bfcfb79b84f676770e49418e153b2af78f1b06c4e854b6291d152cfa8602"},"schema_version":"1.0","source":{"id":"2412.02893","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.02893","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"arxiv_version","alias_value":"2412.02893v1","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.02893","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_12","alias_value":"RKGRAZWWPKOH","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_16","alias_value":"RKGRAZWWPKOH6X3G","created_at":"2026-07-05T09:44:21Z"},{"alias_kind":"pith_short_8","alias_value":"RKGRAZWW","created_at":"2026-07-05T09:44:21Z"}],"graph_snapshots":[{"event_id":"sha256:21619d003a39e6f0cbed14beb5cc62c0c42a5053b5b0a095cc0d375c10e6656e","target":"graph","created_at":"2026-07-05T09:44:21Z","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/2412.02893/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The existing algorithms for identification of neurons responsible for undesired and harmful behaviors do not consider the effects of confounders such as topic of the conversation. In this work, we show that confounders can create spurious correlations and propose a new causal mediation approach that controls the impact of the topic. In experiments with two large language models, we study the localization hypothesis and show that adjusting for the effect of conversation topic, toxicity becomes less localized.","authors_text":"David Heckerman, Milad Fotouhi, Mohammad Taha Bahadori, Oluwaseyi Feyisetan, Payman Arabshahi","cross_cats":["cs.AI","cs.LG","stat.AP","stat.ME"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-03T22:58:21Z","title":"Removing Spurious Correlation from Neural Network Interpretations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.02893","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":"sha256:0068cf3900b08b5dcc5b4ed1ebcaebff49d6241a401af0ff984483f8d85b8b92","target":"record","created_at":"2026-07-05T09:44:21Z","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":"a172c9a7789826c11321e47d30d72ba5c0ea8ae9a7333155c85f2c1a6f83c8b8","cross_cats_sorted":["cs.AI","cs.LG","stat.AP","stat.ME"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-03T22:58:21Z","title_canon_sha256":"b913bfcfb79b84f676770e49418e153b2af78f1b06c4e854b6291d152cfa8602"},"schema_version":"1.0","source":{"id":"2412.02893","kind":"arxiv","version":1}},"canonical_sha256":"8a8d1066d67a9c7f5f66976a341ad5a1b264e931fb1de4a5b89605127a463da5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8a8d1066d67a9c7f5f66976a341ad5a1b264e931fb1de4a5b89605127a463da5","first_computed_at":"2026-07-05T09:44:21.563983Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:44:21.563983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"d+Bs6eKfdTu644azRU/kd7TjxisSkm5y2ug+NQxp8L6E61arCcpvl6gcOAUlYHceU7zsD94gkFLPGJxVRdIxDA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:44:21.564546Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.02893","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0068cf3900b08b5dcc5b4ed1ebcaebff49d6241a401af0ff984483f8d85b8b92","sha256:21619d003a39e6f0cbed14beb5cc62c0c42a5053b5b0a095cc0d375c10e6656e"],"state_sha256":"1a225c41d060d544a0fbb0a5826b0de786aa0e07168e71b7f08f4896f03e5946"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OX6iLjfim4+J/rN/vr9vgqmupNKf6n4GyCYcGiBXng0flm2UR1h4oRLzTQCkQI3KfMKL0u+bQbmwCdDrrgmuAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T09:33:02.129921Z","bundle_sha256":"586d539c5d5d29dd318fe3984522e0523f632bab48b24c0981201fb7372d7438"}}