{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KM5YXNLUJQS6RSOQUTPU2ZFG6U","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":"ba5769e8625109101590cd839b78154c2bb8a90d017253a686310e8f107267e2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-06-22T02:45:01Z","title_canon_sha256":"e64ac311b02bbdf7bc294909a039d632a62f2971ad6583cb92efc16b30a8e29a"},"schema_version":"1.0","source":{"id":"2606.22783","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.22783","created_at":"2026-06-23T02:13:54Z"},{"alias_kind":"arxiv_version","alias_value":"2606.22783v1","created_at":"2026-06-23T02:13:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22783","created_at":"2026-06-23T02:13:54Z"},{"alias_kind":"pith_short_12","alias_value":"KM5YXNLUJQS6","created_at":"2026-06-23T02:13:54Z"},{"alias_kind":"pith_short_16","alias_value":"KM5YXNLUJQS6RSOQ","created_at":"2026-06-23T02:13:54Z"},{"alias_kind":"pith_short_8","alias_value":"KM5YXNLU","created_at":"2026-06-23T02:13:54Z"}],"graph_snapshots":[{"event_id":"sha256:1bb483e4e5002457a72aa9bb13dfd27270f3f8ce691373583c264534b5f62ef8","target":"graph","created_at":"2026-06-23T02:13:54Z","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/2606.22783/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating the exhaustive search capabilities of large language models (LLMs) is plagued by a fundamental paradox: verifying completeness requires complete ground truth, yet high-entropy enumeration tasks make such ground truth impossible for humans to create. This causes benchmarks to systematically penalize models for outperforming their human annotators. Despite rapid progress in web-search and deep research agents -- which now issue hundreds of queries, traverse diverse sites, and synthesize long reports -- evaluation still largely relies on partially annotated answer sets, LLM-based judge","authors_text":"Juntao Wu, Ke Wang, Ruizhi Qiao, Shuai Pang, Wei Wen, Xianting Huang, Xing Sun","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-06-22T02:45:01Z","title":"Breaking the Evaluation Paradox: Evaluating High-Entropy Search with Computationally Irreducible Constraints"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22783","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":"sha256:7cf6edfa335d830408a18bde42c9fd4d86c82e3ab9a135a517c7e252d01a4435","target":"record","created_at":"2026-06-23T02:13:54Z","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":"ba5769e8625109101590cd839b78154c2bb8a90d017253a686310e8f107267e2","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2026-06-22T02:45:01Z","title_canon_sha256":"e64ac311b02bbdf7bc294909a039d632a62f2971ad6583cb92efc16b30a8e29a"},"schema_version":"1.0","source":{"id":"2606.22783","kind":"arxiv","version":1}},"canonical_sha256":"533b8bb5744c25e8c9d0a4df4d64a6f5333a4ec43dae3732a8bb6c28dc9ff6ea","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"533b8bb5744c25e8c9d0a4df4d64a6f5333a4ec43dae3732a8bb6c28dc9ff6ea","first_computed_at":"2026-06-23T02:13:54.425504Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-23T02:13:54.425504Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"HbkRCDWBRkzGItS4KDkBB8x2b9/LTMOcU+62OfIUil2Z+5WaxL/z7STIEgbLkRcie/8peXCm5jdrVEm0hdDzBw==","signature_status":"signed_v1","signed_at":"2026-06-23T02:13:54.425907Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.22783","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7cf6edfa335d830408a18bde42c9fd4d86c82e3ab9a135a517c7e252d01a4435","sha256:1bb483e4e5002457a72aa9bb13dfd27270f3f8ce691373583c264534b5f62ef8"],"state_sha256":"b9a4270a86cff06020cd0f8af6dcfa3c9b91243d0efc71a4885866a5bd8898ed"}