analiza aiqbench
MADE introduces a structured agentic framework that transforms noisy, metric-heavy benchmark data into actionable, fine-grained diagnostic insights across 26 languages. By decomposing post-evaluation analysis into specific steps like instance-level inspection and cultural reflection, the system significantly outperforms standard baselines in report quality and expert preference. This approach effectively bridges the gap between raw evaluation scores and practical guidance for model selection and cross-cultural deployment.
MADE: Beyond Scoring via a Multilingual Agentic Diagnosing Engine for Fine-Grained Evaluation Insights
arXiv:2606.07020v1 Announce Type: new Abstract: 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 grounded report synthesis. MADE is paired with an expert-led 54-query and 15-language diagnostic set, evaluated on top of a large-scale multilingual evaluation substrate (33 model families, 11 benchmarks, 26 languages, 34 cultures, 8.66M evaluation records). Experiments show that MADE outperforms the strongest shared baseline by 47% in diagnosis report quality and is preferred by human multilingual experts in 87.9% of pairwise comparisons. Applied with multilingual experts, MADE further surfaces four actionable findings on deployment, iteration, and cross-cultural pitfalls, turning benchmark score tables into model-selection and remediation guidance.
Abstract: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 grounded report synthesis. MADE is paired with an expert-led 54-query and 15-language diagnostic set, evaluated on top of a large-scale multilingual evaluation substrate (33 model families, 11 benchmarks, 26 languages, 34 cultures, 8.66M evaluation records). Experiments show that MADE outperforms the strongest shared baseline by 47% in diagnosis report quality and is preferred by human multilingual experts in 87.9% of pairwise comparisons. Applied with multilingual experts, MADE further surfaces four actionable findings on deployment, iteration, and cross-cultural pitfalls, turning benchmark score tables into model-selection and remediation guidance.
Computation and Language (cs.CL)
arXiv:2606.07020 [cs.CL]
(or arXiv:2606.07020v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2606.07020 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 181 cuvinte
Subiecte conexe
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · 13 oct. 2026
[AINews] High Taste Testers Ep. 1
Latent Space · 17 oct. 2026
[AINews] Muse Spark 1.3 matches GPT-5.6-Sol, confirming Meta Superintelligence as the newest Frontier Lab, >90% discount for training
Latent Space · 3 sept. 2026
Building trust in AI agents: A production-ready evaluation framework with Amazon Bedrock AgentCore
AWS · 16 oct. 2026