Tag
This paper proposes Judge-LS, a protocol to evaluate whether LLM-as-a-judge models are invariant to language switching between English and Chinese. It finds that switching languages causes 10.7-14.4% preference flips and that judges achieve their highest accuracy in English.
Introduces PoQ-Judge, a multi-architecture evaluation framework with reference-free judge models (TextCNN, MiniLM, DeBERTa) for cost-aware Proof-of-Quality in decentralized LLM inference, achieving high correlation with ground-truth proxies while eliminating the need for reference answers.