{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:LIUNAH6NJHXJHR4NH6VX7VMATS","short_pith_number":"pith:LIUNAH6N","canonical_record":{"source":{"id":"2502.06867","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-08T04:27:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4198290f1b151038215c123bb4722c9e610ed3cc2eedbb1b5267fc6668a121d7","abstract_canon_sha256":"303fce607e2bec3a806c3d9068e5ba597302612ae76e5d04e2015beae03d3dac"},"schema_version":"1.0"},"canonical_sha256":"5a28d01fcd49ee93c78d3fab7fd5809c8624676faffa3494d938a222fdb4c3a6","source":{"kind":"arxiv","id":"2502.06867","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06867","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06867v1","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06867","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_12","alias_value":"LIUNAH6NJHXJ","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_16","alias_value":"LIUNAH6NJHXJHR4N","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_8","alias_value":"LIUNAH6N","created_at":"2026-07-05T10:12:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:LIUNAH6NJHXJHR4NH6VX7VMATS","target":"record","payload":{"canonical_record":{"source":{"id":"2502.06867","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-08T04:27:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4198290f1b151038215c123bb4722c9e610ed3cc2eedbb1b5267fc6668a121d7","abstract_canon_sha256":"303fce607e2bec3a806c3d9068e5ba597302612ae76e5d04e2015beae03d3dac"},"schema_version":"1.0"},"canonical_sha256":"5a28d01fcd49ee93c78d3fab7fd5809c8624676faffa3494d938a222fdb4c3a6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:12:29.109026Z","signature_b64":"mjjtzb8WUK9AW0TWmfs6Yl3wklHLwRRNrftuo1Z1WCmxz5zGk62lTLuU19OIIjpICcehsQua/VLvUffgAxziAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5a28d01fcd49ee93c78d3fab7fd5809c8624676faffa3494d938a222fdb4c3a6","last_reissued_at":"2026-07-05T10:12:29.108559Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:12:29.108559Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.06867","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":"2026-07-05T10:12:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VPXhtEjAPxd0q+cVt+qwzlqkH/zaO9UKMwhr/aALhJqzvEuDxV4S53ZT4yt80ZbKJEXUJXPDGGgDsdGJHHdtDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T20:24:28.049117Z"},"content_sha256":"0923b446bac09df54413f4fca55f029b6bec361eb7986f1f46232eca6823cfcf","schema_version":"1.0","event_id":"sha256:0923b446bac09df54413f4fca55f029b6bec361eb7986f1f46232eca6823cfcf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:LIUNAH6NJHXJHR4NH6VX7VMATS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Forbidden Science: Dual-Use AI Challenge Benchmark and Scientific Refusal Tests","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"David Noever, Forrest McKee","submitted_at":"2025-02-08T04:27:33Z","abstract_excerpt":"The development of robust safety benchmarks for large language models requires open, reproducible datasets that can measure both appropriate refusal of harmful content and potential over-restriction of legitimate scientific discourse. We present an open-source dataset and testing framework for evaluating LLM safety mechanisms across mainly controlled substance queries, analyzing four major models' responses to systematically varied prompts. Our results reveal distinct safety profiles: Claude-3.5-sonnet demonstrated the most conservative approach with 73% refusals and 27% allowances, while Mist"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06867","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/2502.06867/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":"2026-07-05T10:12:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9iqrW40vOwr6i70PoyAsScqIU9pdX2tX2jSR78PA4JwxrMKKtzMEJMB3iTvW97hXoYY3uTHU6Fni64/Uj3JjAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T20:24:28.049602Z"},"content_sha256":"f352c2796a02f7e895c24a5ea18e43d04923f58cd050c45bd5e9602569f5bdbf","schema_version":"1.0","event_id":"sha256:f352c2796a02f7e895c24a5ea18e43d04923f58cd050c45bd5e9602569f5bdbf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/bundle.json","state_url":"https://pith.science/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/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":"2026-08-04T20:24:28Z","links":{"resolver":"https://pith.science/pith/LIUNAH6NJHXJHR4NH6VX7VMATS","bundle":"https://pith.science/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/bundle.json","state":"https://pith.science/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LIUNAH6NJHXJHR4NH6VX7VMATS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:LIUNAH6NJHXJHR4NH6VX7VMATS","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":"303fce607e2bec3a806c3d9068e5ba597302612ae76e5d04e2015beae03d3dac","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-08T04:27:33Z","title_canon_sha256":"4198290f1b151038215c123bb4722c9e610ed3cc2eedbb1b5267fc6668a121d7"},"schema_version":"1.0","source":{"id":"2502.06867","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06867","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06867v1","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06867","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_12","alias_value":"LIUNAH6NJHXJ","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_16","alias_value":"LIUNAH6NJHXJHR4N","created_at":"2026-07-05T10:12:29Z"},{"alias_kind":"pith_short_8","alias_value":"LIUNAH6N","created_at":"2026-07-05T10:12:29Z"}],"graph_snapshots":[{"event_id":"sha256:f352c2796a02f7e895c24a5ea18e43d04923f58cd050c45bd5e9602569f5bdbf","target":"graph","created_at":"2026-07-05T10:12:29Z","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/2502.06867/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The development of robust safety benchmarks for large language models requires open, reproducible datasets that can measure both appropriate refusal of harmful content and potential over-restriction of legitimate scientific discourse. We present an open-source dataset and testing framework for evaluating LLM safety mechanisms across mainly controlled substance queries, analyzing four major models' responses to systematically varied prompts. Our results reveal distinct safety profiles: Claude-3.5-sonnet demonstrated the most conservative approach with 73% refusals and 27% allowances, while Mist","authors_text":"David Noever, Forrest McKee","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-08T04:27:33Z","title":"Forbidden Science: Dual-Use AI Challenge Benchmark and Scientific Refusal Tests"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06867","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":"sha256:0923b446bac09df54413f4fca55f029b6bec361eb7986f1f46232eca6823cfcf","target":"record","created_at":"2026-07-05T10:12:29Z","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":"303fce607e2bec3a806c3d9068e5ba597302612ae76e5d04e2015beae03d3dac","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-08T04:27:33Z","title_canon_sha256":"4198290f1b151038215c123bb4722c9e610ed3cc2eedbb1b5267fc6668a121d7"},"schema_version":"1.0","source":{"id":"2502.06867","kind":"arxiv","version":1}},"canonical_sha256":"5a28d01fcd49ee93c78d3fab7fd5809c8624676faffa3494d938a222fdb4c3a6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5a28d01fcd49ee93c78d3fab7fd5809c8624676faffa3494d938a222fdb4c3a6","first_computed_at":"2026-07-05T10:12:29.108559Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:12:29.108559Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mjjtzb8WUK9AW0TWmfs6Yl3wklHLwRRNrftuo1Z1WCmxz5zGk62lTLuU19OIIjpICcehsQua/VLvUffgAxziAA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:12:29.109026Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.06867","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0923b446bac09df54413f4fca55f029b6bec361eb7986f1f46232eca6823cfcf","sha256:f352c2796a02f7e895c24a5ea18e43d04923f58cd050c45bd5e9602569f5bdbf"],"state_sha256":"fb72b19bb455f56ed7c14023b050a54f33478a868fb2d1924db194c20d7208b8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wnh0b5l9JdmCjonccqvuXV8UikGxsM1K9iMLR4GVavXlyTOkejTSUN8OxoYWJdUzo/TASecmdBBIge4XfKT5Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T20:24:28.052709Z","bundle_sha256":"9e47348ea574d8b8e39de8148761863b296356625ec1d399d338bd61f6b91e32"}}