{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:NKIOL3VWQLF5GJERKQPTB6XR55","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":"a74fa01587040fadcb768e1396108809297364cea3761d149500ebfd39e5952d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-14T07:07:13Z","title_canon_sha256":"133a1ffbc93acca1c61eac356087fe88b2e80cde46f2d8db55cc16908b6d3b16"},"schema_version":"1.0","source":{"id":"2405.08366","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.08366","created_at":"2026-07-05T08:20:54Z"},{"alias_kind":"arxiv_version","alias_value":"2405.08366v3","created_at":"2026-07-05T08:20:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.08366","created_at":"2026-07-05T08:20:54Z"},{"alias_kind":"pith_short_12","alias_value":"NKIOL3VWQLF5","created_at":"2026-07-05T08:20:54Z"},{"alias_kind":"pith_short_16","alias_value":"NKIOL3VWQLF5GJER","created_at":"2026-07-05T08:20:54Z"},{"alias_kind":"pith_short_8","alias_value":"NKIOL3VW","created_at":"2026-07-05T08:20:54Z"}],"graph_snapshots":[{"event_id":"sha256:7c937725d030374f843b89719ca7aebe75ddb8957240498d2a7d2625266c1a7f","target":"graph","created_at":"2026-07-05T08:20:54Z","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/2405.08366/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Disentangling model activations into meaningful features is a central problem in interpretability. However, the absence of ground-truth for these features in realistic scenarios makes validating recent approaches, such as sparse dictionary learning, elusive. To address this challenge, we propose a framework for evaluating feature dictionaries in the context of specific tasks, by comparing them against \\emph{supervised} feature dictionaries. First, we demonstrate that supervised dictionaries achieve excellent approximation, control, and interpretability of model computations on the task. Second","authors_text":"Aleksandar Makelov, George Lange, Neel Nanda","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-14T07:07:13Z","title":"Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Control"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.08366","kind":"arxiv","version":3},"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":"sha256:7868b2a4edcbea3a146d6e67cfea9be1dc0b99f60de3fcf682822d5b6b3030c4","target":"record","created_at":"2026-07-05T08:20:54Z","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":"a74fa01587040fadcb768e1396108809297364cea3761d149500ebfd39e5952d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-14T07:07:13Z","title_canon_sha256":"133a1ffbc93acca1c61eac356087fe88b2e80cde46f2d8db55cc16908b6d3b16"},"schema_version":"1.0","source":{"id":"2405.08366","kind":"arxiv","version":3}},"canonical_sha256":"6a90e5eeb682cbd32491541f30faf1ef477d3d3005c31d1aed28163097b64b3f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6a90e5eeb682cbd32491541f30faf1ef477d3d3005c31d1aed28163097b64b3f","first_computed_at":"2026-07-05T08:20:54.645739Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:20:54.645739Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Bbf7lj3K+FqH8kdY7B5Mc4gx/alNyG0QR8T1t8FZA0ZzWxF79KqLUc+1tYNfk+r4eXiOEyP+lO8HVEOXYJW0CQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:20:54.646228Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.08366","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7868b2a4edcbea3a146d6e67cfea9be1dc0b99f60de3fcf682822d5b6b3030c4","sha256:7c937725d030374f843b89719ca7aebe75ddb8957240498d2a7d2625266c1a7f"],"state_sha256":"cd900d3915bbdaec9edd6b43af47619dfc811d9c552dccade0c69ff7ade91926"}