{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:C3IMHWEUY52WMADHPHRCQTCLQS","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":"f8fb10b7e43f8a929cf17884b5dd85fd7d2af8ff970ff5d207dc63b14c1e00c2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-05T08:09:46Z","title_canon_sha256":"cbf8fc0b006ae69bc3bf96303b34f32b797ddd0de42c8c61cb7d6bcd6a79742a"},"schema_version":"1.0","source":{"id":"2606.07020","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.07020","created_at":"2026-06-08T01:04:41Z"},{"alias_kind":"arxiv_version","alias_value":"2606.07020v1","created_at":"2026-06-08T01:04:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.07020","created_at":"2026-06-08T01:04:41Z"},{"alias_kind":"pith_short_12","alias_value":"C3IMHWEUY52W","created_at":"2026-06-08T01:04:41Z"},{"alias_kind":"pith_short_16","alias_value":"C3IMHWEUY52WMADH","created_at":"2026-06-08T01:04:41Z"},{"alias_kind":"pith_short_8","alias_value":"C3IMHWEU","created_at":"2026-06-08T01:04:41Z"}],"graph_snapshots":[{"event_id":"sha256:4fee8c7e53d90bb056d2c54779051a655b2b8c81c0ab4fa4e78d58f5f7074d4e","target":"graph","created_at":"2026-06-08T01:04:41Z","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/2606.07020/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multilingual and multicultural benchmarks now cover dozens of languages and model families, but the resulting score landscapes remain metric-rich and insight-poor, necessitating fine-grained multilingual post-evaluation diagnosis. However, single LLMs and open-ended agents are easily swamped by the long, noisy diagnostic input, and no reusable taxonomy exists for it. To address this, we propose MADE, a Multilingual Agentic Diagnosing Engine that decomposes post-evaluation analysis into planning, aggregate analysis, instance-level case inspection, multilingual and cultural reflection, and groun","authors_text":"Cheng Qian, Chen Liu, Chenxin Liu, Chunguang Zhao, Daimeng Wei, Liqun Deng, Li Zhang, Miao Zhang, Minggui He, Shimin Tao, Xiaojun Meng, Yilun Liu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-05T08:09:46Z","title":"MADE: Beyond Scoring via a Multilingual Agentic Diagnosing Engine for Fine-Grained Evaluation Insights"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.07020","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":"sha256:4bd8bae3a846f0fe2f5baa737ff66abc77ce94507f4084d7c31baf41445b96f4","target":"record","created_at":"2026-06-08T01:04:41Z","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":"f8fb10b7e43f8a929cf17884b5dd85fd7d2af8ff970ff5d207dc63b14c1e00c2","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-06-05T08:09:46Z","title_canon_sha256":"cbf8fc0b006ae69bc3bf96303b34f32b797ddd0de42c8c61cb7d6bcd6a79742a"},"schema_version":"1.0","source":{"id":"2606.07020","kind":"arxiv","version":1}},"canonical_sha256":"16d0c3d894c77566006779e2284c4b848a996dcb146231e90365e35af2ce803f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"16d0c3d894c77566006779e2284c4b848a996dcb146231e90365e35af2ce803f","first_computed_at":"2026-06-08T01:04:41.713067Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-08T01:04:41.713067Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JyaTVF2FOZZM8qTp/HE1H5JPGLzLhgkf4wIEOka5zeB3zOFkKxsHHRTntd82zjkHMwcQHIFfulUE8YKqQ99NBw==","signature_status":"signed_v1","signed_at":"2026-06-08T01:04:41.713942Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.07020","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4bd8bae3a846f0fe2f5baa737ff66abc77ce94507f4084d7c31baf41445b96f4","sha256:4fee8c7e53d90bb056d2c54779051a655b2b8c81c0ab4fa4e78d58f5f7074d4e"],"state_sha256":"abfcd3f4494715869600dd455d02b58dae33fc7011774c91d595f002bbaf17a5"}