{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:XM7PW25HTNXLYYX5VM6J3PGTKW","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":"96782dc38b3cf9120ab90349b2340e58b835fd88eacb531c3ac73e4aaa6bd684","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MA","submitted_at":"2025-06-03T16:33:47Z","title_canon_sha256":"baeae7aee83866003aa246d176d34546fd97f4df57e82a5d6cf60897c6cef186"},"schema_version":"1.0","source":{"id":"2506.03053","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.03053","created_at":"2026-07-05T11:34:51Z"},{"alias_kind":"arxiv_version","alias_value":"2506.03053v2","created_at":"2026-07-05T11:34:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.03053","created_at":"2026-07-05T11:34:51Z"},{"alias_kind":"pith_short_12","alias_value":"XM7PW25HTNXL","created_at":"2026-07-05T11:34:51Z"},{"alias_kind":"pith_short_16","alias_value":"XM7PW25HTNXLYYX5","created_at":"2026-07-05T11:34:51Z"},{"alias_kind":"pith_short_8","alias_value":"XM7PW25H","created_at":"2026-07-05T11:34:51Z"}],"graph_snapshots":[{"event_id":"sha256:0679d247f2fe8dafac910c92c696b7623359125b29b59f0880eecf0aa5339b58","target":"graph","created_at":"2026-07-05T11:34:51Z","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/2506.03053/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Traditional AI safety evaluations on isolated LLMs are insufficient as multi-agent AI ensembles become prevalent, introducing novel emergent risks. This paper introduces the Multi-Agent Emergent Behavior Evaluation (MAEBE) framework to systematically assess such risks. Using MAEBE with the Greatest Good Benchmark (and a novel double-inversion question technique), we demonstrate that: (1) LLM moral preferences, particularly for Instrumental Harm, are surprisingly brittle and shift significantly with question framing, both in single agents and ensembles. (2) The moral reasoning of LLM ensembles ","authors_text":"Martin Leitgab (Independent Researcher), Ram Potham (Independent Researcher), Sinem Erisken (Independent Researcher), Timothy Gothard (Independent Researcher)","cross_cats":["cs.AI","cs.CL","cs.CY","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MA","submitted_at":"2025-06-03T16:33:47Z","title":"MAEBE: Multi-Agent Emergent Behavior Framework"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.03053","kind":"arxiv","version":2},"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":"sha256:3e265884262cc8229bbbd44fb28bf870e5ec91927cc566b0cc63ddad7829f49a","target":"record","created_at":"2026-07-05T11:34:51Z","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":"96782dc38b3cf9120ab90349b2340e58b835fd88eacb531c3ac73e4aaa6bd684","cross_cats_sorted":["cs.AI","cs.CL","cs.CY","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MA","submitted_at":"2025-06-03T16:33:47Z","title_canon_sha256":"baeae7aee83866003aa246d176d34546fd97f4df57e82a5d6cf60897c6cef186"},"schema_version":"1.0","source":{"id":"2506.03053","kind":"arxiv","version":2}},"canonical_sha256":"bb3efb6ba79b6ebc62fdab3c9dbcd3558528b990cf736c4c75c4e44b778a2554","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bb3efb6ba79b6ebc62fdab3c9dbcd3558528b990cf736c4c75c4e44b778a2554","first_computed_at":"2026-07-05T11:34:51.892205Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:34:51.892205Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VmZs8UJEXznEu9V0CkpNGW8q68ifQ/BOzYLK7z/ECNKWPciYR44WwZV/XuNnXQvPeQ2CaMVecsKxdEHyr3/FDA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:34:51.892731Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.03053","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3e265884262cc8229bbbd44fb28bf870e5ec91927cc566b0cc63ddad7829f49a","sha256:0679d247f2fe8dafac910c92c696b7623359125b29b59f0880eecf0aa5339b58"],"state_sha256":"7ce6a735028418673b46c6731c4296faf720350710c18f87fcdd419cba5ae94b"}