{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:VIRGBTQSNNUQIMDVGANX2IRHEA","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":"ba2dd3fab323aab38bd3e1d8d8442edf1ed68cdde3a026609b15954809bc2d68","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-26T15:44:58Z","title_canon_sha256":"c1f99394efce96d99df7e1d42f24f1e85e1614ced57328de4db2230c24d17cf6"},"schema_version":"1.0","source":{"id":"2403.17806","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.17806","created_at":"2026-07-05T08:43:54Z"},{"alias_kind":"arxiv_version","alias_value":"2403.17806v2","created_at":"2026-07-05T08:43:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.17806","created_at":"2026-07-05T08:43:54Z"},{"alias_kind":"pith_short_12","alias_value":"VIRGBTQSNNUQ","created_at":"2026-07-05T08:43:54Z"},{"alias_kind":"pith_short_16","alias_value":"VIRGBTQSNNUQIMDV","created_at":"2026-07-05T08:43:54Z"},{"alias_kind":"pith_short_8","alias_value":"VIRGBTQS","created_at":"2026-07-05T08:43:54Z"}],"graph_snapshots":[{"event_id":"sha256:75e8ab5572630bdd4ca996f213ad7ad156010f094827eecf0a3475c34a341051","target":"graph","created_at":"2026-07-05T08:43:54Z","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/2403.17806/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Many recent language model (LM) interpretability studies have adopted the circuits framework, which aims to find the minimal computational subgraph, or circuit, that explains LM behavior on a given task. Most studies determine which edges belong in a LM's circuit by performing causal interventions on each edge independently, but this scales poorly with model size. Edge attribution patching (EAP), gradient-based approximation to interventions, has emerged as a scalable but imperfect solution to this problem. In this paper, we introduce a new method - EAP with integrated gradients (EAP-IG) - tha","authors_text":"Michael Hanna, Sandro Pezzelle, Yonatan Belinkov","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-26T15:44:58Z","title":"Have Faith in Faithfulness: Going Beyond Circuit Overlap When Finding Model Mechanisms"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.17806","kind":"arxiv","version":2},"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":"sha256:23ae218389e87492965bbb4e9806b8f790a672e6ea2fcd0cac01878791c904a6","target":"record","created_at":"2026-07-05T08:43:54Z","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":"ba2dd3fab323aab38bd3e1d8d8442edf1ed68cdde3a026609b15954809bc2d68","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-03-26T15:44:58Z","title_canon_sha256":"c1f99394efce96d99df7e1d42f24f1e85e1614ced57328de4db2230c24d17cf6"},"schema_version":"1.0","source":{"id":"2403.17806","kind":"arxiv","version":2}},"canonical_sha256":"aa2260ce126b69043075301b7d2227201c7cff72f5c181f9e61d48b4f97a233a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"aa2260ce126b69043075301b7d2227201c7cff72f5c181f9e61d48b4f97a233a","first_computed_at":"2026-07-05T08:43:54.796395Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:43:54.796395Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RBcy7Kh+/tRYwf9tbAMcytDDRGgxD2EH8pDintu6ZchxOBcPadpPWXK7fVaznuY2ENCKsDcplhPsNmHl8ccSCg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:43:54.796812Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.17806","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:23ae218389e87492965bbb4e9806b8f790a672e6ea2fcd0cac01878791c904a6","sha256:75e8ab5572630bdd4ca996f213ad7ad156010f094827eecf0a3475c34a341051"],"state_sha256":"ee27d2a75db0948d263d675b0429cf012f52260ed565949fdc55a843324001fa"}