{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:7X35R26IJOKXUK6ODWUMLD4P6J","short_pith_number":"pith:7X35R26I","canonical_record":{"source":{"id":"2606.20408","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-18T15:57:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"570114d7bbbdbd5f4043ad224f26f611479b491469f76792e4a59ab9aa90ee9d","abstract_canon_sha256":"378118fcf708d4ff07624e1fb27ce547ccba7bad6abcf9fddc7a05f66a8a28ea"},"schema_version":"1.0"},"canonical_sha256":"fdf7d8ebc84b957a2bce1da8c58f8ff275b121b1fb477ffa5412491c270f4ca7","source":{"kind":"arxiv","id":"2606.20408","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20408","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20408v1","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20408","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_12","alias_value":"7X35R26IJOKX","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_16","alias_value":"7X35R26IJOKXUK6O","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_8","alias_value":"7X35R26I","created_at":"2026-06-19T16:13:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:7X35R26IJOKXUK6ODWUMLD4P6J","target":"record","payload":{"canonical_record":{"source":{"id":"2606.20408","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-18T15:57:53Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"570114d7bbbdbd5f4043ad224f26f611479b491469f76792e4a59ab9aa90ee9d","abstract_canon_sha256":"378118fcf708d4ff07624e1fb27ce547ccba7bad6abcf9fddc7a05f66a8a28ea"},"schema_version":"1.0"},"canonical_sha256":"fdf7d8ebc84b957a2bce1da8c58f8ff275b121b1fb477ffa5412491c270f4ca7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:13:11.780656Z","signature_b64":"j4oV6AVPVDk6s3yT/rEb3B/92iXpVCoyOhH5oex82B0ZYV1FSpkMTveR/8DW3u+MsWzAPJt5pVmDzsMzoM/TAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fdf7d8ebc84b957a2bce1da8c58f8ff275b121b1fb477ffa5412491c270f4ca7","last_reissued_at":"2026-06-19T16:13:11.780280Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:13:11.780280Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.20408","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":"2026-06-19T16:13:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6cQi+E2b99KKT9+Q6viI1MA7etEw7zj7WeEfHrb0gRff9XAn2sHb2qcC4HgwGX34CgJ+xBsdAHyyFt73s1z1CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T08:41:45.151597Z"},"content_sha256":"b9eccb8eba30f957597216394eb6d121406a265f788b6a1cf6b23116cdca0523","schema_version":"1.0","event_id":"sha256:b9eccb8eba30f957597216394eb6d121406a265f788b6a1cf6b23116cdca0523"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:7X35R26IJOKXUK6ODWUMLD4P6J","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLM agent safety, multi-turn red-teaming, jailbreak benchmarks, adversarial robustness, safety-critical systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Bokyeong Kim, Dasol Choi, Hanwool Lee, Haon Park, Seung Geun Kim","submitted_at":"2026-06-18T15:57:53Z","abstract_excerpt":"Large language model (LLM) agents are increasingly proposed as supervisory components for safety-critical systems, yet their robustness under sustained, adaptive adversarial pressure remains poorly characterized. We present NRT-Bench, a benchmark for multi-turn red-teaming of LLM agents acting as operators of a safety-critical system, instantiated in a simulated nuclear power plant control room. A five-role operator team, each backed by a configurable LLM, runs a plant governed by six critical safety functions (CSFs), while adversaries inject messages over four channels in bounded multi-turn s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20408","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/2606.20408/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":"2026-06-19T16:13:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iALjnhxJSwNq/hh6XckMJr2Ef1T8dJA78ieoaLieb7JrCsa8t6zgTZ05S7AX9UN0oXV1ietU9KCxuUbEvZpsCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T08:41:45.151979Z"},"content_sha256":"1a815acbecd0991f62d3c7537af854a0050e5d56be53d127c0063e8d26f262f5","schema_version":"1.0","event_id":"sha256:1a815acbecd0991f62d3c7537af854a0050e5d56be53d127c0063e8d26f262f5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7X35R26IJOKXUK6ODWUMLD4P6J/bundle.json","state_url":"https://pith.science/pith/7X35R26IJOKXUK6ODWUMLD4P6J/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7X35R26IJOKXUK6ODWUMLD4P6J/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":"2026-06-30T08:41:45Z","links":{"resolver":"https://pith.science/pith/7X35R26IJOKXUK6ODWUMLD4P6J","bundle":"https://pith.science/pith/7X35R26IJOKXUK6ODWUMLD4P6J/bundle.json","state":"https://pith.science/pith/7X35R26IJOKXUK6ODWUMLD4P6J/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7X35R26IJOKXUK6ODWUMLD4P6J/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7X35R26IJOKXUK6ODWUMLD4P6J","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":"378118fcf708d4ff07624e1fb27ce547ccba7bad6abcf9fddc7a05f66a8a28ea","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-18T15:57:53Z","title_canon_sha256":"570114d7bbbdbd5f4043ad224f26f611479b491469f76792e4a59ab9aa90ee9d"},"schema_version":"1.0","source":{"id":"2606.20408","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20408","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20408v1","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20408","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_12","alias_value":"7X35R26IJOKX","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_16","alias_value":"7X35R26IJOKXUK6O","created_at":"2026-06-19T16:13:11Z"},{"alias_kind":"pith_short_8","alias_value":"7X35R26I","created_at":"2026-06-19T16:13:11Z"}],"graph_snapshots":[{"event_id":"sha256:1a815acbecd0991f62d3c7537af854a0050e5d56be53d127c0063e8d26f262f5","target":"graph","created_at":"2026-06-19T16:13:11Z","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/2606.20408/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language model (LLM) agents are increasingly proposed as supervisory components for safety-critical systems, yet their robustness under sustained, adaptive adversarial pressure remains poorly characterized. We present NRT-Bench, a benchmark for multi-turn red-teaming of LLM agents acting as operators of a safety-critical system, instantiated in a simulated nuclear power plant control room. A five-role operator team, each backed by a configurable LLM, runs a plant governed by six critical safety functions (CSFs), while adversaries inject messages over four channels in bounded multi-turn s","authors_text":"Bokyeong Kim, Dasol Choi, Hanwool Lee, Haon Park, Seung Geun Kim","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-18T15:57:53Z","title":"LLM agent safety, multi-turn red-teaming, jailbreak benchmarks, adversarial robustness, safety-critical systems"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20408","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:b9eccb8eba30f957597216394eb6d121406a265f788b6a1cf6b23116cdca0523","target":"record","created_at":"2026-06-19T16:13:11Z","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":"378118fcf708d4ff07624e1fb27ce547ccba7bad6abcf9fddc7a05f66a8a28ea","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2026-06-18T15:57:53Z","title_canon_sha256":"570114d7bbbdbd5f4043ad224f26f611479b491469f76792e4a59ab9aa90ee9d"},"schema_version":"1.0","source":{"id":"2606.20408","kind":"arxiv","version":1}},"canonical_sha256":"fdf7d8ebc84b957a2bce1da8c58f8ff275b121b1fb477ffa5412491c270f4ca7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"fdf7d8ebc84b957a2bce1da8c58f8ff275b121b1fb477ffa5412491c270f4ca7","first_computed_at":"2026-06-19T16:13:11.780280Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:13:11.780280Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"j4oV6AVPVDk6s3yT/rEb3B/92iXpVCoyOhH5oex82B0ZYV1FSpkMTveR/8DW3u+MsWzAPJt5pVmDzsMzoM/TAA==","signature_status":"signed_v1","signed_at":"2026-06-19T16:13:11.780656Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.20408","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b9eccb8eba30f957597216394eb6d121406a265f788b6a1cf6b23116cdca0523","sha256:1a815acbecd0991f62d3c7537af854a0050e5d56be53d127c0063e8d26f262f5"],"state_sha256":"c70a8c5ff153d9fa70a402dfc4cd9b59f3783705c513709cc3a93283064f5e4e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"srLjpMzLPRw6lV1+vmXbNExZuPAP0LM4VeGO5xQX3Vben1MBy6nsuGwObb7n/3mlg6baJGYK4WbIg5MwGrngCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-30T08:41:45.154287Z","bundle_sha256":"53f5ee3cc6e3acb7b534851371fa20435c8497b93056e6f08316379075028fbb"}}