{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:BSV6WLVJNG33CBTC353TI6Q7BL","short_pith_number":"pith:BSV6WLVJ","canonical_record":{"source":{"id":"2411.07122","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:51:39Z","cross_cats_sorted":[],"title_canon_sha256":"7b832f2369891f38dc5484301273ebd6479d29e12661e51035a3853d67390782","abstract_canon_sha256":"dedc6a3f200b243c64d557efe9b750d828fa31a197a2ef0e9cbe12813d14eeda"},"schema_version":"1.0"},"canonical_sha256":"0cabeb2ea969b7b10662df77347a1f0adc54fb7d13fe88138e7db5dd1f123bd1","source":{"kind":"arxiv","id":"2411.07122","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.07122","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"arxiv_version","alias_value":"2411.07122v2","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.07122","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_12","alias_value":"BSV6WLVJNG33","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_16","alias_value":"BSV6WLVJNG33CBTC","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_8","alias_value":"BSV6WLVJ","created_at":"2026-07-05T09:44:54Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:BSV6WLVJNG33CBTC353TI6Q7BL","target":"record","payload":{"canonical_record":{"source":{"id":"2411.07122","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:51:39Z","cross_cats_sorted":[],"title_canon_sha256":"7b832f2369891f38dc5484301273ebd6479d29e12661e51035a3853d67390782","abstract_canon_sha256":"dedc6a3f200b243c64d557efe9b750d828fa31a197a2ef0e9cbe12813d14eeda"},"schema_version":"1.0"},"canonical_sha256":"0cabeb2ea969b7b10662df77347a1f0adc54fb7d13fe88138e7db5dd1f123bd1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:44:54.532049Z","signature_b64":"0izzaiO84q/57vQUgcEjJ3+hLukAPMUhv/J6Z5+qQFDik2DmO+nnmu1wYsQI6JX1Opw6xnNqnMcVOkc+QVC9BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0cabeb2ea969b7b10662df77347a1f0adc54fb7d13fe88138e7db5dd1f123bd1","last_reissued_at":"2026-07-05T09:44:54.531553Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:44:54.531553Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.07122","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:44:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CurnJ43pu9dEEsuXZ18IPomdKYSemlPuBCZN8wlj2K9xDXL7soVdcsDy+DbEQZrP5iS+qN0L6vIKYQhw3WFUDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T23:56:51.725094Z"},"content_sha256":"6acb42b2e2a401053f069fcd4a320a4c0ac93ed60930b5f8755152c3aab93bfc","schema_version":"1.0","event_id":"sha256:6acb42b2e2a401053f069fcd4a320a4c0ac93ed60930b5f8755152c3aab93bfc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:BSV6WLVJNG33CBTC353TI6Q7BL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SCAR: Sparse Conditioned Autoencoders for Concept Detection and Steering in LLMs","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bj\\\"orn Deiseroth, Felix Friedrich, Kristian Kersting, Manuel Brack, Patrick Schramowski, Ruben H\\\"arle","submitted_at":"2024-11-11T16:51:39Z","abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable capabilities in generating human-like text, but their output may not be aligned with the user or even produce harmful content. This paper presents a novel approach to detect and steer concepts such as toxicity before generation. We introduce the Sparse Conditioned Autoencoder (SCAR), a single trained module that extends the otherwise untouched LLM. SCAR ensures full steerability, towards and away from concepts (e.g., toxic content), without compromising the quality of the model's text generation on standard evaluation benchmarks. We dem"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.07122","kind":"arxiv","version":2},"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/2411.07122/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":"2026-07-05T09:44:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Aikd+GSY0bySY104sfjnYAQHHqvuaSzmBgcD3dAy+ddfOu91ZaqyJoCIwBBao6xgc1Ip1j5dMjwH0Bfp6FWjAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T23:56:51.725677Z"},"content_sha256":"b3a9d73afb94f361abad7581b922a2df589e653813952ff48df234952dfcbaf6","schema_version":"1.0","event_id":"sha256:b3a9d73afb94f361abad7581b922a2df589e653813952ff48df234952dfcbaf6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BSV6WLVJNG33CBTC353TI6Q7BL/bundle.json","state_url":"https://pith.science/pith/BSV6WLVJNG33CBTC353TI6Q7BL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BSV6WLVJNG33CBTC353TI6Q7BL/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":"2026-08-09T23:56:51Z","links":{"resolver":"https://pith.science/pith/BSV6WLVJNG33CBTC353TI6Q7BL","bundle":"https://pith.science/pith/BSV6WLVJNG33CBTC353TI6Q7BL/bundle.json","state":"https://pith.science/pith/BSV6WLVJNG33CBTC353TI6Q7BL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BSV6WLVJNG33CBTC353TI6Q7BL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:BSV6WLVJNG33CBTC353TI6Q7BL","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":"dedc6a3f200b243c64d557efe9b750d828fa31a197a2ef0e9cbe12813d14eeda","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:51:39Z","title_canon_sha256":"7b832f2369891f38dc5484301273ebd6479d29e12661e51035a3853d67390782"},"schema_version":"1.0","source":{"id":"2411.07122","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.07122","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"arxiv_version","alias_value":"2411.07122v2","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.07122","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_12","alias_value":"BSV6WLVJNG33","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_16","alias_value":"BSV6WLVJNG33CBTC","created_at":"2026-07-05T09:44:54Z"},{"alias_kind":"pith_short_8","alias_value":"BSV6WLVJ","created_at":"2026-07-05T09:44:54Z"}],"graph_snapshots":[{"event_id":"sha256:b3a9d73afb94f361abad7581b922a2df589e653813952ff48df234952dfcbaf6","target":"graph","created_at":"2026-07-05T09:44:54Z","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/2411.07122/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable capabilities in generating human-like text, but their output may not be aligned with the user or even produce harmful content. This paper presents a novel approach to detect and steer concepts such as toxicity before generation. We introduce the Sparse Conditioned Autoencoder (SCAR), a single trained module that extends the otherwise untouched LLM. SCAR ensures full steerability, towards and away from concepts (e.g., toxic content), without compromising the quality of the model's text generation on standard evaluation benchmarks. We dem","authors_text":"Bj\\\"orn Deiseroth, Felix Friedrich, Kristian Kersting, Manuel Brack, Patrick Schramowski, Ruben H\\\"arle","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:51:39Z","title":"SCAR: Sparse Conditioned Autoencoders for Concept Detection and Steering in LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.07122","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":"sha256:6acb42b2e2a401053f069fcd4a320a4c0ac93ed60930b5f8755152c3aab93bfc","target":"record","created_at":"2026-07-05T09:44:54Z","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":"dedc6a3f200b243c64d557efe9b750d828fa31a197a2ef0e9cbe12813d14eeda","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-11-11T16:51:39Z","title_canon_sha256":"7b832f2369891f38dc5484301273ebd6479d29e12661e51035a3853d67390782"},"schema_version":"1.0","source":{"id":"2411.07122","kind":"arxiv","version":2}},"canonical_sha256":"0cabeb2ea969b7b10662df77347a1f0adc54fb7d13fe88138e7db5dd1f123bd1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0cabeb2ea969b7b10662df77347a1f0adc54fb7d13fe88138e7db5dd1f123bd1","first_computed_at":"2026-07-05T09:44:54.531553Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:44:54.531553Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0izzaiO84q/57vQUgcEjJ3+hLukAPMUhv/J6Z5+qQFDik2DmO+nnmu1wYsQI6JX1Opw6xnNqnMcVOkc+QVC9BQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:44:54.532049Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.07122","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6acb42b2e2a401053f069fcd4a320a4c0ac93ed60930b5f8755152c3aab93bfc","sha256:b3a9d73afb94f361abad7581b922a2df589e653813952ff48df234952dfcbaf6"],"state_sha256":"b91f61f479ebf5b73a6577355df64f16f056e1595b8257aeed345008728379de"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Z8BiZ5A6GrkwXYaLppT16obGJTmEgGy/1GVAmJWHAPrK9j388rUHUH6jae60IsOkBjoxYOQsxZBpDbr7T9w+DA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T23:56:51.732618Z","bundle_sha256":"0424db50eb6ed2b8b6bcaa03f82e73b1665ff21d086c3f63739eaac210018332"}}