{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:5RC57MVHXYFM2UWJVVXEBE4UGZ","short_pith_number":"pith:5RC57MVH","canonical_record":{"source":{"id":"2311.04235","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-11-06T08:50:29Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"0d3ecb9f59edc928b360b39735e66e4c78e8520115e1e4f6ca225882f97e2df6","abstract_canon_sha256":"7ea6a1a247193d32b7cc09a5ffca79e5fe5ab842d7c322b2af1d13f10c037a8e"},"schema_version":"1.0"},"canonical_sha256":"ec45dfb2a7be0acd52c9ad6e40939436477eba652faf35650e417beae00b704b","source":{"kind":"arxiv","id":"2311.04235","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.04235","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"arxiv_version","alias_value":"2311.04235v3","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.04235","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_12","alias_value":"5RC57MVHXYFM","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_16","alias_value":"5RC57MVHXYFM2UWJ","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_8","alias_value":"5RC57MVH","created_at":"2026-07-05T07:53:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:5RC57MVHXYFM2UWJVVXEBE4UGZ","target":"record","payload":{"canonical_record":{"source":{"id":"2311.04235","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-11-06T08:50:29Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"0d3ecb9f59edc928b360b39735e66e4c78e8520115e1e4f6ca225882f97e2df6","abstract_canon_sha256":"7ea6a1a247193d32b7cc09a5ffca79e5fe5ab842d7c322b2af1d13f10c037a8e"},"schema_version":"1.0"},"canonical_sha256":"ec45dfb2a7be0acd52c9ad6e40939436477eba652faf35650e417beae00b704b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:53:48.939769Z","signature_b64":"4M/DXwVeJSpAJdRxFyB/a9PAtb+X9KzeM1E1ClknDSUoSDSJsiD22hY0EAL36sMKIJdJ2I5rgDQVshLJNQWcDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ec45dfb2a7be0acd52c9ad6e40939436477eba652faf35650e417beae00b704b","last_reissued_at":"2026-07-05T07:53:48.939286Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:53:48.939286Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.04235","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:53:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R9iHkyrIB8MGWIah0EYYqGCUUdQPRFUN9RbLW7Kax1Ivmo4EgCYnWKSJ8FqLt0PdhgvhngU7N31Q11zwXHMJCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T00:01:50.692218Z"},"content_sha256":"93a20378922c6b710c1703e30ec3e10d5f95fa9b518551381d35f0a366e412c8","schema_version":"1.0","event_id":"sha256:93a20378922c6b710c1703e30ec3e10d5f95fa9b518551381d35f0a366e412c8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:5RC57MVHXYFM2UWJVVXEBE4UGZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can LLMs Follow Simple Rules?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Basel Alomair, Dan Hendrycks, David Karamardian, David Wagner, Lulwa Aljeraisy, Norman Mu, Sarah Chen, Sizhe Chen, Zifan Wang","submitted_at":"2023-11-06T08:50:29Z","abstract_excerpt":"As Large Language Models (LLMs) are deployed with increasing real-world responsibilities, it is important to be able to specify and constrain the behavior of these systems in a reliable manner. Model developers may wish to set explicit rules for the model, such as \"do not generate abusive content\", but these may be circumvented by jailbreaking techniques. Existing evaluations of adversarial attacks and defenses on LLMs generally require either expensive manual review or unreliable heuristic checks. To address this issue, we propose Rule-following Language Evaluation Scenarios (RuLES), a progra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.04235","kind":"arxiv","version":3},"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/2311.04235/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":"2026-07-05T07:53:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aVgGFWeK85XBR3yn+tRBzH1EPhxPLuBAdocFRorZFyol+mXkTEOgwkB8Rwzo6+6gXYYh3NLlx99HoTP1kw7IDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T00:01:50.693039Z"},"content_sha256":"7f640d184ec1cefff0ba3b7435c186ea474f2b2169e58c4f43c158c59550b323","schema_version":"1.0","event_id":"sha256:7f640d184ec1cefff0ba3b7435c186ea474f2b2169e58c4f43c158c59550b323"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/bundle.json","state_url":"https://pith.science/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/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":"2026-08-05T00:01:50Z","links":{"resolver":"https://pith.science/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ","bundle":"https://pith.science/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/bundle.json","state":"https://pith.science/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5RC57MVHXYFM2UWJVVXEBE4UGZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:5RC57MVHXYFM2UWJVVXEBE4UGZ","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":"7ea6a1a247193d32b7cc09a5ffca79e5fe5ab842d7c322b2af1d13f10c037a8e","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-11-06T08:50:29Z","title_canon_sha256":"0d3ecb9f59edc928b360b39735e66e4c78e8520115e1e4f6ca225882f97e2df6"},"schema_version":"1.0","source":{"id":"2311.04235","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.04235","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"arxiv_version","alias_value":"2311.04235v3","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.04235","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_12","alias_value":"5RC57MVHXYFM","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_16","alias_value":"5RC57MVHXYFM2UWJ","created_at":"2026-07-05T07:53:48Z"},{"alias_kind":"pith_short_8","alias_value":"5RC57MVH","created_at":"2026-07-05T07:53:48Z"}],"graph_snapshots":[{"event_id":"sha256:7f640d184ec1cefff0ba3b7435c186ea474f2b2169e58c4f43c158c59550b323","target":"graph","created_at":"2026-07-05T07:53:48Z","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/2311.04235/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As Large Language Models (LLMs) are deployed with increasing real-world responsibilities, it is important to be able to specify and constrain the behavior of these systems in a reliable manner. Model developers may wish to set explicit rules for the model, such as \"do not generate abusive content\", but these may be circumvented by jailbreaking techniques. Existing evaluations of adversarial attacks and defenses on LLMs generally require either expensive manual review or unreliable heuristic checks. To address this issue, we propose Rule-following Language Evaluation Scenarios (RuLES), a progra","authors_text":"Basel Alomair, Dan Hendrycks, David Karamardian, David Wagner, Lulwa Aljeraisy, Norman Mu, Sarah Chen, Sizhe Chen, Zifan Wang","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-11-06T08:50:29Z","title":"Can LLMs Follow Simple Rules?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.04235","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":"sha256:93a20378922c6b710c1703e30ec3e10d5f95fa9b518551381d35f0a366e412c8","target":"record","created_at":"2026-07-05T07:53:48Z","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":"7ea6a1a247193d32b7cc09a5ffca79e5fe5ab842d7c322b2af1d13f10c037a8e","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2023-11-06T08:50:29Z","title_canon_sha256":"0d3ecb9f59edc928b360b39735e66e4c78e8520115e1e4f6ca225882f97e2df6"},"schema_version":"1.0","source":{"id":"2311.04235","kind":"arxiv","version":3}},"canonical_sha256":"ec45dfb2a7be0acd52c9ad6e40939436477eba652faf35650e417beae00b704b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ec45dfb2a7be0acd52c9ad6e40939436477eba652faf35650e417beae00b704b","first_computed_at":"2026-07-05T07:53:48.939286Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:53:48.939286Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4M/DXwVeJSpAJdRxFyB/a9PAtb+X9KzeM1E1ClknDSUoSDSJsiD22hY0EAL36sMKIJdJ2I5rgDQVshLJNQWcDA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:53:48.939769Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.04235","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:93a20378922c6b710c1703e30ec3e10d5f95fa9b518551381d35f0a366e412c8","sha256:7f640d184ec1cefff0ba3b7435c186ea474f2b2169e58c4f43c158c59550b323"],"state_sha256":"76f17f6c6154107b8369252640375ec1ed31ffb0bb5be9c1da23f1aa562277e3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xVeAUHgCSYKuqiYG6rHk+kDuHAkxF0dTUrh1mwo8cqW0o3MHzQj8jhp+Q9b2JLUOt6YMItJ3LFSHe1+QLIhbAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T00:01:50.699821Z","bundle_sha256":"db32fe9e14c6fdb2b97b1680884693763227bf7d5215b6a0681f1c700f02f37f"}}