{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HPG2X7MDOHOPILIZYYV2TQUOYQ","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":"d5aab075b7923f04b43f00ab6befe45e7620b24c9d2ff9f38affbcd4306fb8de","cross_cats_sorted":["cs.AI","cs.NE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-09-05T18:00:37Z","title_canon_sha256":"4e1ea02ec6a0b8c2dfdfabddca1cb05a7341b2d36a09e698a8c9a7f3eb55ae8c"},"schema_version":"1.0","source":{"id":"2409.04478","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.04478","created_at":"2026-07-05T09:04:12Z"},{"alias_kind":"arxiv_version","alias_value":"2409.04478v1","created_at":"2026-07-05T09:04:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.04478","created_at":"2026-07-05T09:04:12Z"},{"alias_kind":"pith_short_12","alias_value":"HPG2X7MDOHOP","created_at":"2026-07-05T09:04:12Z"},{"alias_kind":"pith_short_16","alias_value":"HPG2X7MDOHOPILIZ","created_at":"2026-07-05T09:04:12Z"},{"alias_kind":"pith_short_8","alias_value":"HPG2X7MD","created_at":"2026-07-05T09:04:12Z"}],"graph_snapshots":[{"event_id":"sha256:a7e3c7d91f578d87ab889af24e03924aa38c662ab1bda9ea2766fb52ea0dfb83","target":"graph","created_at":"2026-07-05T09:04:12Z","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/2409.04478/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"A popular new method in mechanistic interpretability is to train high-dimensional sparse autoencoders (SAEs) on neuron activations and use SAE features as the atomic units of analysis. However, the body of evidence on whether SAE feature spaces are useful for causal analysis is underdeveloped. In this work, we use the RAVEL benchmark to evaluate whether SAEs trained on hidden representations of GPT-2 small have sets of features that separately mediate knowledge of which country a city is in and which continent it is in. We evaluate four open-source SAEs for GPT-2 small against each other, with","authors_text":"Atticus Geiger, Maheep Chaudhary","cross_cats":["cs.AI","cs.NE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-09-05T18:00:37Z","title":"Evaluating Open-Source Sparse Autoencoders on Disentangling Factual Knowledge in GPT-2 Small"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.04478","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":"sha256:525a89c634fb0e0ccbfe7adf3e949d54777d048664d02360a504614fa5d7754c","target":"record","created_at":"2026-07-05T09:04:12Z","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":"d5aab075b7923f04b43f00ab6befe45e7620b24c9d2ff9f38affbcd4306fb8de","cross_cats_sorted":["cs.AI","cs.NE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-09-05T18:00:37Z","title_canon_sha256":"4e1ea02ec6a0b8c2dfdfabddca1cb05a7341b2d36a09e698a8c9a7f3eb55ae8c"},"schema_version":"1.0","source":{"id":"2409.04478","kind":"arxiv","version":1}},"canonical_sha256":"3bcdabfd8371dcf42d19c62ba9c28ec40ff610eeb5b4e30622f4da337f59c170","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3bcdabfd8371dcf42d19c62ba9c28ec40ff610eeb5b4e30622f4da337f59c170","first_computed_at":"2026-07-05T09:04:12.676080Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:04:12.676080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4M38dxzCkvjGoaaD5f70wYbxpHsBarsG6SCBmzRSvPAoLYIKHdN41u8OzfSWwM4s2wGAU295x0tFV8taQZHkBw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:04:12.676501Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.04478","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:525a89c634fb0e0ccbfe7adf3e949d54777d048664d02360a504614fa5d7754c","sha256:a7e3c7d91f578d87ab889af24e03924aa38c662ab1bda9ea2766fb52ea0dfb83"],"state_sha256":"2ea7b02b13a6775619f2bbc384a3e9b531c21fe18bc5b4a0de21a5779fdf6c03"}