{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:C4OLGEL43KNQYY3EK64LMLCWJF","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":"0555dcd221d01f7fb61661ba2a3baeff3eca475d9482ac181c671ecd9cecb21f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-28T23:35:28Z","title_canon_sha256":"093ec8654a4f03ad45b28e5fd35d9e867c429854338ca172bb7cb90baa16c361"},"schema_version":"1.0","source":{"id":"2606.29646","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.29646","created_at":"2026-06-30T02:17:28Z"},{"alias_kind":"arxiv_version","alias_value":"2606.29646v1","created_at":"2026-06-30T02:17:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29646","created_at":"2026-06-30T02:17:28Z"},{"alias_kind":"pith_short_12","alias_value":"C4OLGEL43KNQ","created_at":"2026-06-30T02:17:28Z"},{"alias_kind":"pith_short_16","alias_value":"C4OLGEL43KNQYY3E","created_at":"2026-06-30T02:17:28Z"},{"alias_kind":"pith_short_8","alias_value":"C4OLGEL4","created_at":"2026-06-30T02:17:28Z"}],"graph_snapshots":[{"event_id":"sha256:30b011179d411429d9728c99b73ffdc189f6d79a34b950e2bf722966bc4b0c8c","target":"graph","created_at":"2026-06-30T02:17:28Z","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/2606.29646/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Sleeper agents are the canonical model organism of deception: models trained to behave normally but to emit an unsafe behaviour on a specific trigger. Eliciting that behaviour without knowing the trigger has not been studied systematically. We study fuzzing: injecting Gaussian noise into a model's weights or residual-stream activations and checking whether the perturbed outputs reveal the behaviour. On 6 backdoored models (7B-13B) we compare both forms of fuzzing head-to-head against temperature-sampling baselines. Fuzzing elicits the hidden behaviour more often than temperature sampling on 4 ","authors_text":"Lakshmi Babu-Saheer, Mohammed Abu Baker","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-28T23:35:28Z","title":"Fuzzing Large Language Models to Elicit Hidden Behaviours"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29646","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":"sha256:3be6ce0b66a224311167ecfc8ca44516fc1d50252f9a04595001156407dc4ae0","target":"record","created_at":"2026-06-30T02:17:28Z","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":"0555dcd221d01f7fb61661ba2a3baeff3eca475d9482ac181c671ecd9cecb21f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-28T23:35:28Z","title_canon_sha256":"093ec8654a4f03ad45b28e5fd35d9e867c429854338ca172bb7cb90baa16c361"},"schema_version":"1.0","source":{"id":"2606.29646","kind":"arxiv","version":1}},"canonical_sha256":"171cb3117cda9b0c636457b8b62c5649506faa6ac6bd3ba5f530c21d661deced","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"171cb3117cda9b0c636457b8b62c5649506faa6ac6bd3ba5f530c21d661deced","first_computed_at":"2026-06-30T02:17:28.777502Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-30T02:17:28.777502Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Wy1pmD6YAmHKq+qPWGY55vP+RIeMJpBzeT6+nd74H7NQ6z2NfNaGrpw/rueXkVCmT/Ch9K1+QaZIl1Kyu/S6BQ==","signature_status":"signed_v1","signed_at":"2026-06-30T02:17:28.778229Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.29646","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3be6ce0b66a224311167ecfc8ca44516fc1d50252f9a04595001156407dc4ae0","sha256:30b011179d411429d9728c99b73ffdc189f6d79a34b950e2bf722966bc4b0c8c"],"state_sha256":"791f1277053b9dbedc248dc85e1015bf0f28bce1f2ba3995ff08f36fa5139bcc"}