{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:U5LFUISKAI2XO7KVPO2PEHKCCX","short_pith_number":"pith:U5LFUISK","schema_version":"1.0","canonical_sha256":"a7565a224a0235777d557bb4f21d4215f4120e516f710f7d927af3278aa3cb36","source":{"kind":"arxiv","id":"2605.21849","version":1},"attestation_state":"computed","paper":{"title":"Geometry-Adaptive Explainer for Faithful Dictionary-Based Interpretability under Distribution Shift","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Andrew Lee, Heedong Kim, Kyungwoo Song, Sungjun Lim","submitted_at":"2026-05-21T00:46:01Z","abstract_excerpt":"Mechanistic interpretability aims to explain a model's behavior by identifying causally responsible internal structures. Dictionary-based explainers such as sparse autoencoders and transcoders are a primary tool, but their faithfulness under out-of-distribution (OOD) shift has received little systematic attention. We show that distribution shift rotates the subspace that the model actively uses, misaligning the explainer's dictionary trained on in-distribution (ID) activations. We formalize this misalignment as the faithfulness gap, a geometric distance between the ID dictionary and the OOD-ac"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.21849","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-21T00:46:01Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6dd1e73f9fb7d82f3bf599907d43cd65124c0d0ec52930028b6357fc1dfa4933","abstract_canon_sha256":"82ed5b0d60f7f9427b799ed678e9d5f33beeeb4861034b21abc5ab6c631b4222"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T01:04:11.035141Z","signature_b64":"neUzUYrYc2KsAwPY7jYxWF82gsqY5IsOZrh00MGdjHWnqzaoXbl7D/X4MNC0dYYQvqP2V8wMvJZxCtPXc2+CCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a7565a224a0235777d557bb4f21d4215f4120e516f710f7d927af3278aa3cb36","last_reissued_at":"2026-05-22T01:04:11.034394Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T01:04:11.034394Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Geometry-Adaptive Explainer for Faithful Dictionary-Based Interpretability under Distribution Shift","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Andrew Lee, Heedong Kim, Kyungwoo Song, Sungjun Lim","submitted_at":"2026-05-21T00:46:01Z","abstract_excerpt":"Mechanistic interpretability aims to explain a model's behavior by identifying causally responsible internal structures. Dictionary-based explainers such as sparse autoencoders and transcoders are a primary tool, but their faithfulness under out-of-distribution (OOD) shift has received little systematic attention. We show that distribution shift rotates the subspace that the model actively uses, misaligning the explainer's dictionary trained on in-distribution (ID) activations. We formalize this misalignment as the faithfulness gap, a geometric distance between the ID dictionary and the OOD-ac"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21849","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/2605.21849/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.21849","created_at":"2026-05-22T01:04:11.034500+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.21849v1","created_at":"2026-05-22T01:04:11.034500+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21849","created_at":"2026-05-22T01:04:11.034500+00:00"},{"alias_kind":"pith_short_12","alias_value":"U5LFUISKAI2X","created_at":"2026-05-22T01:04:11.034500+00:00"},{"alias_kind":"pith_short_16","alias_value":"U5LFUISKAI2XO7KV","created_at":"2026-05-22T01:04:11.034500+00:00"},{"alias_kind":"pith_short_8","alias_value":"U5LFUISK","created_at":"2026-05-22T01:04:11.034500+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX","json":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX.json","graph_json":"https://pith.science/api/pith-number/U5LFUISKAI2XO7KVPO2PEHKCCX/graph.json","events_json":"https://pith.science/api/pith-number/U5LFUISKAI2XO7KVPO2PEHKCCX/events.json","paper":"https://pith.science/paper/U5LFUISK"},"agent_actions":{"view_html":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX","download_json":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX.json","view_paper":"https://pith.science/paper/U5LFUISK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.21849&json=true","fetch_graph":"https://pith.science/api/pith-number/U5LFUISKAI2XO7KVPO2PEHKCCX/graph.json","fetch_events":"https://pith.science/api/pith-number/U5LFUISKAI2XO7KVPO2PEHKCCX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX/action/storage_attestation","attest_author":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX/action/author_attestation","sign_citation":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX/action/citation_signature","submit_replication":"https://pith.science/pith/U5LFUISKAI2XO7KVPO2PEHKCCX/action/replication_record"}},"created_at":"2026-05-22T01:04:11.034500+00:00","updated_at":"2026-05-22T01:04:11.034500+00:00"}