{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:OZN6LXXSWBW3V57R2PMK2W6RQK","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":"d0370297a8a660fe74dec084015f326d741e9b2dcb37a24a31208797300c04f3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T16:45:18Z","title_canon_sha256":"73ae18cb0d5bd9c43153c8a961072380c381c279c096130be15c84cb7dd6f22c"},"schema_version":"1.0","source":{"id":"2502.06655","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.06655","created_at":"2026-07-05T11:01:31Z"},{"alias_kind":"arxiv_version","alias_value":"2502.06655v2","created_at":"2026-07-05T11:01:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06655","created_at":"2026-07-05T11:01:31Z"},{"alias_kind":"pith_short_12","alias_value":"OZN6LXXSWBW3","created_at":"2026-07-05T11:01:31Z"},{"alias_kind":"pith_short_16","alias_value":"OZN6LXXSWBW3V57R","created_at":"2026-07-05T11:01:31Z"},{"alias_kind":"pith_short_8","alias_value":"OZN6LXXS","created_at":"2026-07-05T11:01:31Z"}],"graph_snapshots":[{"event_id":"sha256:0c16c2f3ba8cbffebf857224a277c3a584d7727b899c374e63101048f90d7689","target":"graph","created_at":"2026-07-05T11:01:31Z","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/2502.06655/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Benchmark contamination has become a significant concern in the LLM evaluation community. Previous Agents-as-an-Evaluator address this issue by involving agents in the generation of questions. Despite their success, the biases in Agents-as-an-Evaluator methods remain largely unexplored. In this paper, we present a theoretical formulation of evaluation bias, providing valuable insights into designing unbiased evaluation protocols. Furthermore, we identify two type of bias in Agents-as-an-Evaluator through carefully designed probing tasks on a minimal Agents-as-an-Evaluator setup. To address the","authors_text":"Di Xie, Jiang Zhu, Jian Tian, Liang Ma, Meilin Chen, Weijie Chen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T16:45:18Z","title":"Unbiased Evaluation of Large Language Models from a Causal Perspective"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06655","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:d89864b01e9869bfe9c0bc12624a5fedad509a59e8b0d84ba0c61120fd11a01e","target":"record","created_at":"2026-07-05T11:01:31Z","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":"d0370297a8a660fe74dec084015f326d741e9b2dcb37a24a31208797300c04f3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-10T16:45:18Z","title_canon_sha256":"73ae18cb0d5bd9c43153c8a961072380c381c279c096130be15c84cb7dd6f22c"},"schema_version":"1.0","source":{"id":"2502.06655","kind":"arxiv","version":2}},"canonical_sha256":"765be5def2b06dbaf7f1d3d8ad5bd182b4b3d8d611874f9841e658b2200f77ce","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"765be5def2b06dbaf7f1d3d8ad5bd182b4b3d8d611874f9841e658b2200f77ce","first_computed_at":"2026-07-05T11:01:31.743209Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:01:31.743209Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eQr28G4oznywH4H8pQnEadYqw9FrqbGVrojMZO2MeyGVUOHYDiusAQCKfxWBUU88JS51tx0a9FpSatlfGEXBBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:01:31.743700Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.06655","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d89864b01e9869bfe9c0bc12624a5fedad509a59e8b0d84ba0c61120fd11a01e","sha256:0c16c2f3ba8cbffebf857224a277c3a584d7727b899c374e63101048f90d7689"],"state_sha256":"62b3630b125d9b67d317aae26d0f15ada95b0f6e845cef9af84f04191a182dca"}