{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:Z7WJ3OD226FX4J25N5SJ3P7FVI","short_pith_number":"pith:Z7WJ3OD2","canonical_record":{"source":{"id":"2303.04381","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-08T05:09:59Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1366cb4cddba8f5294c0198730ae3d2be38a54957072469866d4c39ad089401c","abstract_canon_sha256":"b8796d98b0b16773da5627b8f8526926cc0d13582e3aeae4de300c642780d60c"},"schema_version":"1.0"},"canonical_sha256":"cfec9db87ad78b7e275d6f649dbfe5aa273a1c43d99b5cd47724a65b880b155d","source":{"kind":"arxiv","id":"2303.04381","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2303.04381","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"arxiv_version","alias_value":"2303.04381v1","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.04381","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_12","alias_value":"Z7WJ3OD226FX","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_16","alias_value":"Z7WJ3OD226FX4J25","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_8","alias_value":"Z7WJ3OD2","created_at":"2026-07-05T05:49:20Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:Z7WJ3OD226FX4J25N5SJ3P7FVI","target":"record","payload":{"canonical_record":{"source":{"id":"2303.04381","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-08T05:09:59Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"1366cb4cddba8f5294c0198730ae3d2be38a54957072469866d4c39ad089401c","abstract_canon_sha256":"b8796d98b0b16773da5627b8f8526926cc0d13582e3aeae4de300c642780d60c"},"schema_version":"1.0"},"canonical_sha256":"cfec9db87ad78b7e275d6f649dbfe5aa273a1c43d99b5cd47724a65b880b155d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:49:20.745824Z","signature_b64":"S7UNOGkTTwQkDg4HJtXlFsi58MV5otXi4Plp1fedaGp5OS+FW0W05cfQ6JJbyx0WZxiftlgmzgmkhbMBYTrsBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cfec9db87ad78b7e275d6f649dbfe5aa273a1c43d99b5cd47724a65b880b155d","last_reissued_at":"2026-07-05T05:49:20.745347Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:49:20.745347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2303.04381","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:49:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bzHOTexOHWJd/ungiL2DwsBIFTtCkQAtR/XFkiyiDGTjMyKu+PATPYhftVvfl1QYJ84IaCkdaItch8T6N5UaAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T17:26:02.786203Z"},"content_sha256":"638b3c8d4866785a565f177d958b2fb570cb7cd9db019db2c09040ece51c5c18","schema_version":"1.0","event_id":"sha256:638b3c8d4866785a565f177d958b2fb570cb7cd9db019db2c09040ece51c5c18"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:Z7WJ3OD226FX4J25N5SJ3P7FVI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Automatically Auditing Large Language Models via Discrete Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Aditi Raghunathan, Anca Dragan, Erik Jones, Jacob Steinhardt","submitted_at":"2023-03-08T05:09:59Z","abstract_excerpt":"Auditing large language models for unexpected behaviors is critical to preempt catastrophic deployments, yet remains challenging. In this work, we cast auditing as an optimization problem, where we automatically search for input-output pairs that match a desired target behavior. For example, we might aim to find a non-toxic input that starts with \"Barack Obama\" that a model maps to a toxic output. This optimization problem is difficult to solve as the set of feasible points is sparse, the space is discrete, and the language models we audit are non-linear and high-dimensional. To combat these c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.04381","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/2303.04381/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":"2026-07-05T05:49:20Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AcsIWfAzmLkYKjEo8mRDTm0I5kJeK86HhSisGU9vAjfusAsixkqUJH/GfKZ3S4tdIYyUaRjFw4Xt/yCHwKpyAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T17:26:02.787090Z"},"content_sha256":"80eb0210a30beb6588e081feb79a4f3d3c08c4d07b44a16862cd9886eccdcded","schema_version":"1.0","event_id":"sha256:80eb0210a30beb6588e081feb79a4f3d3c08c4d07b44a16862cd9886eccdcded"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/bundle.json","state_url":"https://pith.science/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/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":"2026-08-06T17:26:02Z","links":{"resolver":"https://pith.science/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI","bundle":"https://pith.science/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/bundle.json","state":"https://pith.science/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Z7WJ3OD226FX4J25N5SJ3P7FVI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:Z7WJ3OD226FX4J25N5SJ3P7FVI","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":"b8796d98b0b16773da5627b8f8526926cc0d13582e3aeae4de300c642780d60c","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-08T05:09:59Z","title_canon_sha256":"1366cb4cddba8f5294c0198730ae3d2be38a54957072469866d4c39ad089401c"},"schema_version":"1.0","source":{"id":"2303.04381","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2303.04381","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"arxiv_version","alias_value":"2303.04381v1","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.04381","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_12","alias_value":"Z7WJ3OD226FX","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_16","alias_value":"Z7WJ3OD226FX4J25","created_at":"2026-07-05T05:49:20Z"},{"alias_kind":"pith_short_8","alias_value":"Z7WJ3OD2","created_at":"2026-07-05T05:49:20Z"}],"graph_snapshots":[{"event_id":"sha256:80eb0210a30beb6588e081feb79a4f3d3c08c4d07b44a16862cd9886eccdcded","target":"graph","created_at":"2026-07-05T05:49:20Z","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/2303.04381/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Auditing large language models for unexpected behaviors is critical to preempt catastrophic deployments, yet remains challenging. In this work, we cast auditing as an optimization problem, where we automatically search for input-output pairs that match a desired target behavior. For example, we might aim to find a non-toxic input that starts with \"Barack Obama\" that a model maps to a toxic output. This optimization problem is difficult to solve as the set of feasible points is sparse, the space is discrete, and the language models we audit are non-linear and high-dimensional. To combat these c","authors_text":"Aditi Raghunathan, Anca Dragan, Erik Jones, Jacob Steinhardt","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-08T05:09:59Z","title":"Automatically Auditing Large Language Models via Discrete Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.04381","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":"sha256:638b3c8d4866785a565f177d958b2fb570cb7cd9db019db2c09040ece51c5c18","target":"record","created_at":"2026-07-05T05:49:20Z","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":"b8796d98b0b16773da5627b8f8526926cc0d13582e3aeae4de300c642780d60c","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-08T05:09:59Z","title_canon_sha256":"1366cb4cddba8f5294c0198730ae3d2be38a54957072469866d4c39ad089401c"},"schema_version":"1.0","source":{"id":"2303.04381","kind":"arxiv","version":1}},"canonical_sha256":"cfec9db87ad78b7e275d6f649dbfe5aa273a1c43d99b5cd47724a65b880b155d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cfec9db87ad78b7e275d6f649dbfe5aa273a1c43d99b5cd47724a65b880b155d","first_computed_at":"2026-07-05T05:49:20.745347Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:49:20.745347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"S7UNOGkTTwQkDg4HJtXlFsi58MV5otXi4Plp1fedaGp5OS+FW0W05cfQ6JJbyx0WZxiftlgmzgmkhbMBYTrsBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T05:49:20.745824Z","signed_message":"canonical_sha256_bytes"},"source_id":"2303.04381","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:638b3c8d4866785a565f177d958b2fb570cb7cd9db019db2c09040ece51c5c18","sha256:80eb0210a30beb6588e081feb79a4f3d3c08c4d07b44a16862cd9886eccdcded"],"state_sha256":"2e57ce18a09fe22bd71c9363e2adc78cc0b449d49ad04e03559c4f5e3c4d3c4e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fijK/9uz0gCfNzNLjCSGOXqUXD92Bqe1uZ3IoVgQcOx5/p8KCWhXj7tReCc81fbBDduLyYhjLT/BW0svN30kDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T17:26:02.794221Z","bundle_sha256":"5ce963c3532f1059fe21ca33159d5a508cb99fb8148e3f461cd53c1bb794a8a9"}}