{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TXE5RINRCSR2ATRENAQKEH6XJZ","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":"0414a704ea17279559225647bea6190ac0ee3aed519286199bad6f87825fbabf","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-30T17:48:35Z","title_canon_sha256":"1e59ca5db1d5817ff0e699fcd987382506aaaeaf1d77b7b02601f2c0a7a45953"},"schema_version":"1.0","source":{"id":"2509.26600","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.26600","created_at":"2026-05-27T01:04:51Z"},{"alias_kind":"arxiv_version","alias_value":"2509.26600v2","created_at":"2026-05-27T01:04:51Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.26600","created_at":"2026-05-27T01:04:51Z"},{"alias_kind":"pith_short_12","alias_value":"TXE5RINRCSR2","created_at":"2026-05-27T01:04:51Z"},{"alias_kind":"pith_short_16","alias_value":"TXE5RINRCSR2ATRE","created_at":"2026-05-27T01:04:51Z"},{"alias_kind":"pith_short_8","alias_value":"TXE5RINR","created_at":"2026-05-27T01:04:51Z"}],"graph_snapshots":[{"event_id":"sha256:9583cc0aee7309db13a34af9743bf419f8778f8d3ebe3f4d364f769b1807fec9","target":"graph","created_at":"2026-05-27T01:04:51Z","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/2509.26600/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As LLMs rapidly saturate existing benchmarks, automated benchmark creation using LLMs (LLM-as-a-benchmark) -- where a model generates test inputs (LLM-as-a-testset) and evaluates outputs (LLM-as-an-evaluator) -- has gained traction as a cheap alternative to human curation. We show that this paradigm has a fundamental problem: LLM-generated benchmarks systematically favor the model that created them. Using machine translation as our primary testbed, we find that self-bias arises from two additive sources, LLM-as-a-testset and LLM-as-an-evaluator, and their combination amplifies the effect. Cruc","authors_text":"Daniel Deutsch, Markus Freitag, Sweta Agrawal, Vil\\'em Zouhar, Wenda Xu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-30T17:48:35Z","title":"When LLMs Benchmark Themselves: Deconstructing Self-Bias in Automated Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.26600","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":"sha256:4f58ec035381e5dab17cab4424430536aae665b323e00d7dcabf873097d4a6ef","target":"record","created_at":"2026-05-27T01:04:51Z","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":"0414a704ea17279559225647bea6190ac0ee3aed519286199bad6f87825fbabf","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-30T17:48:35Z","title_canon_sha256":"1e59ca5db1d5817ff0e699fcd987382506aaaeaf1d77b7b02601f2c0a7a45953"},"schema_version":"1.0","source":{"id":"2509.26600","kind":"arxiv","version":2}},"canonical_sha256":"9dc9d8a1b114a3a04e246820a21fd74e7b8c033daca0c1d2929da24e3d6ad911","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9dc9d8a1b114a3a04e246820a21fd74e7b8c033daca0c1d2929da24e3d6ad911","first_computed_at":"2026-05-27T01:04:51.382102Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-27T01:04:51.382102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uLOHw1wQDU2f2/EejIH/AOdTr3J7Vaq0T+1H5WUbO/XujGclenSY0fnHaPcR9/k3nAiN9MhmyueBGQ3hXTuiAA==","signature_status":"signed_v1","signed_at":"2026-05-27T01:04:51.382963Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.26600","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4f58ec035381e5dab17cab4424430536aae665b323e00d7dcabf873097d4a6ef","sha256:9583cc0aee7309db13a34af9743bf419f8778f8d3ebe3f4d364f769b1807fec9"],"state_sha256":"7b448df34bf27d8590a79aeee8018865ad386d4cb5feecdb9ade6f6234ba0b09"}