{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FCKUV6WWHGSM2BG5PVTFY7MAQM","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":"93759e23e51c821eb9f09a1e359739845ecafcdd2535fe5cae3795a9b45e1f17","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T02:32:35Z","title_canon_sha256":"d7a04be750ec2461b026dd0c0f157e711aa32b81cf0ef4db4bf5f04d495a7c74"},"schema_version":"1.0","source":{"id":"2505.12201","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12201","created_at":"2026-07-05T11:05:00Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12201v1","created_at":"2026-07-05T11:05:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12201","created_at":"2026-07-05T11:05:00Z"},{"alias_kind":"pith_short_12","alias_value":"FCKUV6WWHGSM","created_at":"2026-07-05T11:05:00Z"},{"alias_kind":"pith_short_16","alias_value":"FCKUV6WWHGSM2BG5","created_at":"2026-07-05T11:05:00Z"},{"alias_kind":"pith_short_8","alias_value":"FCKUV6WW","created_at":"2026-07-05T11:05:00Z"}],"graph_snapshots":[{"event_id":"sha256:7f38f9073e2bbb4123335a0d10868d86e0a675f9368043031ff54f6c14a1fd1c","target":"graph","created_at":"2026-07-05T11:05:00Z","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/2505.12201/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"LLM-as-a-Judge has emerged as a popular evaluation strategy, where advanced large language models assess generation results in alignment with human instructions. While these models serve as a promising alternative to human annotators, their reliability in multilingual evaluation remains uncertain. To bridge this gap, we conduct a comprehensive analysis of multilingual LLM-as-a-Judge. Specifically, we evaluate five models from different model families across five diverse tasks involving 25 languages. Our findings reveal that LLMs struggle to achieve consistent judgment results across languages,","authors_text":"Wei Liu, Xiyan Fu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T02:32:35Z","title":"How Reliable is Multilingual LLM-as-a-Judge?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12201","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":"sha256:0de66aebe43f3035f7efc5024a20938f186fddecf009e68a2b81a63acdd90cce","target":"record","created_at":"2026-07-05T11:05:00Z","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":"93759e23e51c821eb9f09a1e359739845ecafcdd2535fe5cae3795a9b45e1f17","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T02:32:35Z","title_canon_sha256":"d7a04be750ec2461b026dd0c0f157e711aa32b81cf0ef4db4bf5f04d495a7c74"},"schema_version":"1.0","source":{"id":"2505.12201","kind":"arxiv","version":1}},"canonical_sha256":"28954afad639a4cd04dd7d665c7d808329cd4796a5d99f8df5a1dc266c27e86d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"28954afad639a4cd04dd7d665c7d808329cd4796a5d99f8df5a1dc266c27e86d","first_computed_at":"2026-07-05T11:05:00.089287Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:00.089287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"gtx/mRUKbTu6P9TA0XRNP/pIHxg6BM8NCHpWe8B/0giC+XytFDQ6FciVlGa3E9qTzyJyAYiFgAnUFuGBTKhQAg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:00.089778Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12201","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0de66aebe43f3035f7efc5024a20938f186fddecf009e68a2b81a63acdd90cce","sha256:7f38f9073e2bbb4123335a0d10868d86e0a675f9368043031ff54f6c14a1fd1c"],"state_sha256":"f341fe2de6fd45433f8ab6c0746fc81846c16f8adfb81ad012d99998eb9d814b"}