{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:LOPMJKJXQ3AX5UZQSKKO4LZP2Z","short_pith_number":"pith:LOPMJKJX","canonical_record":{"source":{"id":"2605.21545","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-20T09:53:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7404e9fc6acb9719626d03f4702e7be4991a91106bafb2f99806ba4b40c46739","abstract_canon_sha256":"608ae7ddc8effe7eb16c97bb16646d39438e56b9a46500b3a2da06c6d475e9e9"},"schema_version":"1.0"},"canonical_sha256":"5b9ec4a93786c17ed3309294ee2f2fd647c08757702178f13f6610b3d4478257","source":{"kind":"arxiv","id":"2605.21545","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21545","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21545v1","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21545","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_12","alias_value":"LOPMJKJXQ3AX","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_16","alias_value":"LOPMJKJXQ3AX5UZQ","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_8","alias_value":"LOPMJKJX","created_at":"2026-05-22T00:02:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:LOPMJKJXQ3AX5UZQSKKO4LZP2Z","target":"record","payload":{"canonical_record":{"source":{"id":"2605.21545","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-20T09:53:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7404e9fc6acb9719626d03f4702e7be4991a91106bafb2f99806ba4b40c46739","abstract_canon_sha256":"608ae7ddc8effe7eb16c97bb16646d39438e56b9a46500b3a2da06c6d475e9e9"},"schema_version":"1.0"},"canonical_sha256":"5b9ec4a93786c17ed3309294ee2f2fd647c08757702178f13f6610b3d4478257","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-22T00:02:28.115208Z","signature_b64":"JO/mb0Qc1/7z3UBv+7EWXVMb1XzO7Br01t6Yaa+GLzqakcfTtffbYCC6mSOr/cW8JBG6gC3wMipumdO++ZojBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b9ec4a93786c17ed3309294ee2f2fd647c08757702178f13f6610b3d4478257","last_reissued_at":"2026-05-22T00:02:28.114760Z","signature_status":"signed_v1","first_computed_at":"2026-05-22T00:02:28.114760Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.21545","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":"2026-05-22T00:02:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SJWrSI/kBpa+l862bj4g4heA8njcbFIneaXrU/qslPXBS0JFXtpgjjkMZGquds6q6JkrwG1VPgIfTSpdGOtKBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T23:57:33.953623Z"},"content_sha256":"738e7b2db2bb546e5c0c1ad6cb3438ac87935fa9d3969af81cd09f2390f184cb","schema_version":"1.0","event_id":"sha256:738e7b2db2bb546e5c0c1ad6cb3438ac87935fa9d3969af81cd09f2390f184cb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:LOPMJKJXQ3AX5UZQSKKO4LZP2Z","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"RefusalBench: Why Refusal Rate Misranks Frontier LLMs on Biological Research Prompts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Aakaash Meduri, Emre Ulgac, Lukas Weidener, Marko Brki\\'c, Mihailo Jovanovi\\'c","submitted_at":"2026-05-20T09:53:31Z","abstract_excerpt":"Frontier large language models are increasingly deployed as orchestration backbones for biological research workflows, yet no shared evidence base exists for comparing their refusal behaviour on legitimate research prompts. RefusalBench, introduced here, is a matched-triple benchmark of 141 prompts in 47 bundles that holds task framing constant while varying only biological risk tier (benign, borderline, dual-use), enabling tier-conditioned comparisons robust to subdomain confounding. A 15-prompt should-refuse positive-control module establishes per-model calibration floors; three models fail "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21545","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/2605.21545/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":"2026-05-22T00:02:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MWr213ySg3RMMuGoxX/TPeIkA0TbQe2VlKtb727M4M47Jtu0UMOlqzuCYBWfqYWWubpej7BDCA9nHq2JnqWBBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T23:57:33.954363Z"},"content_sha256":"1fb408d8370c038ac25793fdd094a48cd4a60cbdfa3fb1cbf5d3f8338e3205cf","schema_version":"1.0","event_id":"sha256:1fb408d8370c038ac25793fdd094a48cd4a60cbdfa3fb1cbf5d3f8338e3205cf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/bundle.json","state_url":"https://pith.science/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/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":"2026-05-22T23:57:33Z","links":{"resolver":"https://pith.science/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z","bundle":"https://pith.science/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/bundle.json","state":"https://pith.science/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LOPMJKJXQ3AX5UZQSKKO4LZP2Z/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:LOPMJKJXQ3AX5UZQSKKO4LZP2Z","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":"608ae7ddc8effe7eb16c97bb16646d39438e56b9a46500b3a2da06c6d475e9e9","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-20T09:53:31Z","title_canon_sha256":"7404e9fc6acb9719626d03f4702e7be4991a91106bafb2f99806ba4b40c46739"},"schema_version":"1.0","source":{"id":"2605.21545","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.21545","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"arxiv_version","alias_value":"2605.21545v1","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21545","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_12","alias_value":"LOPMJKJXQ3AX","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_16","alias_value":"LOPMJKJXQ3AX5UZQ","created_at":"2026-05-22T00:02:28Z"},{"alias_kind":"pith_short_8","alias_value":"LOPMJKJX","created_at":"2026-05-22T00:02:28Z"}],"graph_snapshots":[{"event_id":"sha256:1fb408d8370c038ac25793fdd094a48cd4a60cbdfa3fb1cbf5d3f8338e3205cf","target":"graph","created_at":"2026-05-22T00:02:28Z","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/2605.21545/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Frontier large language models are increasingly deployed as orchestration backbones for biological research workflows, yet no shared evidence base exists for comparing their refusal behaviour on legitimate research prompts. RefusalBench, introduced here, is a matched-triple benchmark of 141 prompts in 47 bundles that holds task framing constant while varying only biological risk tier (benign, borderline, dual-use), enabling tier-conditioned comparisons robust to subdomain confounding. A 15-prompt should-refuse positive-control module establishes per-model calibration floors; three models fail ","authors_text":"Aakaash Meduri, Emre Ulgac, Lukas Weidener, Marko Brki\\'c, Mihailo Jovanovi\\'c","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-20T09:53:31Z","title":"RefusalBench: Why Refusal Rate Misranks Frontier LLMs on Biological Research Prompts"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21545","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":"sha256:738e7b2db2bb546e5c0c1ad6cb3438ac87935fa9d3969af81cd09f2390f184cb","target":"record","created_at":"2026-05-22T00:02:28Z","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":"608ae7ddc8effe7eb16c97bb16646d39438e56b9a46500b3a2da06c6d475e9e9","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-20T09:53:31Z","title_canon_sha256":"7404e9fc6acb9719626d03f4702e7be4991a91106bafb2f99806ba4b40c46739"},"schema_version":"1.0","source":{"id":"2605.21545","kind":"arxiv","version":1}},"canonical_sha256":"5b9ec4a93786c17ed3309294ee2f2fd647c08757702178f13f6610b3d4478257","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5b9ec4a93786c17ed3309294ee2f2fd647c08757702178f13f6610b3d4478257","first_computed_at":"2026-05-22T00:02:28.114760Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-22T00:02:28.114760Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JO/mb0Qc1/7z3UBv+7EWXVMb1XzO7Br01t6Yaa+GLzqakcfTtffbYCC6mSOr/cW8JBG6gC3wMipumdO++ZojBA==","signature_status":"signed_v1","signed_at":"2026-05-22T00:02:28.115208Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.21545","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:738e7b2db2bb546e5c0c1ad6cb3438ac87935fa9d3969af81cd09f2390f184cb","sha256:1fb408d8370c038ac25793fdd094a48cd4a60cbdfa3fb1cbf5d3f8338e3205cf"],"state_sha256":"6398c186c3293c0cd216efb3b1c3bdc767c3fc33257042be49295fae6fe617ca"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Q0ODNvP+Yc3hLJsmj+OCArCqAtmUfNVhEUTWh21LaCTbEt+UxnQ6ridd69Mzay02MV1o2merhlxgwcIJ8LbBDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T23:57:33.957954Z","bundle_sha256":"0a54fe7a24f9f3c4a27f25ca4e43b025c4610bd6e4545614d0f970eb6227663a"}}