{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:KNK3NDX6UQOZJ6PKHPBT7OO3NM","short_pith_number":"pith:KNK3NDX6","canonical_record":{"source":{"id":"2608.04549","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T07:39:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"74ce12b10bb33817f5cd5577439fde12deb5480d5b86f89a75c3ab0768a3b1e0","abstract_canon_sha256":"ecb3ff6a8060a8a17a95d56a55dfa9dcc8fd969d3f9adb9dd6000b1a31aaced7"},"schema_version":"1.0"},"canonical_sha256":"5355b68efea41d94f9ea3bc33fb9db6b2125f6f74d30fd779fc3a69d536d3a91","source":{"kind":"arxiv","id":"2608.04549","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04549","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04549v1","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04549","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_12","alias_value":"KNK3NDX6UQOZ","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_16","alias_value":"KNK3NDX6UQOZJ6PK","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_8","alias_value":"KNK3NDX6","created_at":"2026-08-06T01:35:49Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:KNK3NDX6UQOZJ6PKHPBT7OO3NM","target":"record","payload":{"canonical_record":{"source":{"id":"2608.04549","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T07:39:49Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"74ce12b10bb33817f5cd5577439fde12deb5480d5b86f89a75c3ab0768a3b1e0","abstract_canon_sha256":"ecb3ff6a8060a8a17a95d56a55dfa9dcc8fd969d3f9adb9dd6000b1a31aaced7"},"schema_version":"1.0"},"canonical_sha256":"5355b68efea41d94f9ea3bc33fb9db6b2125f6f74d30fd779fc3a69d536d3a91","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:35:49.133411Z","signature_b64":"/L2TdrTHmZ0iijmX/DKRDQC3B07TcTx4qE4R+H01IloMKsKhUXl7z4w3EFV7GsVzDYwxXIeJ29aEd4GexuieCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5355b68efea41d94f9ea3bc33fb9db6b2125f6f74d30fd779fc3a69d536d3a91","last_reissued_at":"2026-08-06T01:35:49.131806Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:35:49.131806Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.04549","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":"2026-08-06T01:35:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6H+3ePlcFQb9LUNTnBz+81agvaBchS5gZz/5z2lpr+CNmXKTyOgpMHeFnW9YCNPrHKtpPjAByzPGO0WlqJn+AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T10:57:03.399611Z"},"content_sha256":"cacba2a1e188e2b77720f00a6613cee5bd04392e7f94921c6f5ea713e91e32c8","schema_version":"1.0","event_id":"sha256:cacba2a1e188e2b77720f00a6613cee5bd04392e7f94921c6f5ea713e91e32c8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:KNK3NDX6UQOZJ6PKHPBT7OO3NM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"EuroExec: Frontier Language Models Fall Short of Expert Judgment on European Executive Decision Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Amrut Avhad, Danylo Smahliuk, Khaled Denfir, Marcus A. Castro, Pau Arnal","submitted_at":"2026-08-05T07:39:49Z","abstract_excerpt":"Frontier LLMs are increasingly put to use on open-ended complex questions, different in nature from the ones they are typically evaluated on. We dedicate more than 4,000 human expert hours to evaluate a selection of six frontier LLMs on a member of this class of problems: EuroExec, our introduced human expert-based benchmark composed of 413 open-ended long-form European executive tasks authored by 47 vetted domain experts, each question drawn from experience in a real case. Every response is manually evaluated through a multi-attribute rubric, an item-specific checklist of requirements, and a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04549","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/2608.04549/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":"2026-08-06T01:35:49Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fCaqR8A2GXy/nvRMI0F4aUQyk1mrFkQjOWiFI4Rel3kKrtHYC/8zRUH+6IIB/2gMjnowRPYvvBseWGoPuGLfDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-22T10:57:03.400124Z"},"content_sha256":"799a1e97fd7d0bd972e87c38b93bf314b5dd6be8e86443d6235fd79c02244895","schema_version":"1.0","event_id":"sha256:799a1e97fd7d0bd972e87c38b93bf314b5dd6be8e86443d6235fd79c02244895"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/bundle.json","state_url":"https://pith.science/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/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":"2026-08-22T10:57:03Z","links":{"resolver":"https://pith.science/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM","bundle":"https://pith.science/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/bundle.json","state":"https://pith.science/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KNK3NDX6UQOZJ6PKHPBT7OO3NM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KNK3NDX6UQOZJ6PKHPBT7OO3NM","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":"ecb3ff6a8060a8a17a95d56a55dfa9dcc8fd969d3f9adb9dd6000b1a31aaced7","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T07:39:49Z","title_canon_sha256":"74ce12b10bb33817f5cd5577439fde12deb5480d5b86f89a75c3ab0768a3b1e0"},"schema_version":"1.0","source":{"id":"2608.04549","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.04549","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"arxiv_version","alias_value":"2608.04549v1","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04549","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_12","alias_value":"KNK3NDX6UQOZ","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_16","alias_value":"KNK3NDX6UQOZJ6PK","created_at":"2026-08-06T01:35:49Z"},{"alias_kind":"pith_short_8","alias_value":"KNK3NDX6","created_at":"2026-08-06T01:35:49Z"}],"graph_snapshots":[{"event_id":"sha256:799a1e97fd7d0bd972e87c38b93bf314b5dd6be8e86443d6235fd79c02244895","target":"graph","created_at":"2026-08-06T01:35:49Z","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/2608.04549/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Frontier LLMs are increasingly put to use on open-ended complex questions, different in nature from the ones they are typically evaluated on. We dedicate more than 4,000 human expert hours to evaluate a selection of six frontier LLMs on a member of this class of problems: EuroExec, our introduced human expert-based benchmark composed of 413 open-ended long-form European executive tasks authored by 47 vetted domain experts, each question drawn from experience in a real case. Every response is manually evaluated through a multi-attribute rubric, an item-specific checklist of requirements, and a ","authors_text":"Amrut Avhad, Danylo Smahliuk, Khaled Denfir, Marcus A. Castro, Pau Arnal","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T07:39:49Z","title":"EuroExec: Frontier Language Models Fall Short of Expert Judgment on European Executive Decision Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04549","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:cacba2a1e188e2b77720f00a6613cee5bd04392e7f94921c6f5ea713e91e32c8","target":"record","created_at":"2026-08-06T01:35:49Z","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":"ecb3ff6a8060a8a17a95d56a55dfa9dcc8fd969d3f9adb9dd6000b1a31aaced7","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-05T07:39:49Z","title_canon_sha256":"74ce12b10bb33817f5cd5577439fde12deb5480d5b86f89a75c3ab0768a3b1e0"},"schema_version":"1.0","source":{"id":"2608.04549","kind":"arxiv","version":1}},"canonical_sha256":"5355b68efea41d94f9ea3bc33fb9db6b2125f6f74d30fd779fc3a69d536d3a91","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5355b68efea41d94f9ea3bc33fb9db6b2125f6f74d30fd779fc3a69d536d3a91","first_computed_at":"2026-08-06T01:35:49.131806Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-06T01:35:49.131806Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/L2TdrTHmZ0iijmX/DKRDQC3B07TcTx4qE4R+H01IloMKsKhUXl7z4w3EFV7GsVzDYwxXIeJ29aEd4GexuieCA==","signature_status":"signed_v1","signed_at":"2026-08-06T01:35:49.133411Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.04549","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:cacba2a1e188e2b77720f00a6613cee5bd04392e7f94921c6f5ea713e91e32c8","sha256:799a1e97fd7d0bd972e87c38b93bf314b5dd6be8e86443d6235fd79c02244895"],"state_sha256":"b99ad93901a184a554f466e6a92edb0acb6888fcc468c0325e648e3924018c49"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MjwYZ8gGCwZUkalUrLh5no8EAvzf+a5Iq6ncsgUFFB6AlpuAG3BslCe5d880NUJTK1+SlY0u+V2NPCeVvAm4Aw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-22T10:57:03.404454Z","bundle_sha256":"41ce56694d14ec787a75ac1a79fef048fa1e3746cd5c73e3d075140ed7702709"}}