DGrid AI Releases PoQ-Judge Research Paper, Reducing LLM Quality Assessment Costs by 72%

According to ChainCatcher, DGrid AI released its latest research paper "PoQ-Judge" today, introducing a multi-architecture quality assessment framework that eliminates the need for reference answers. The framework achieved a 0.747 correlation with human evaluation scores on held-out test sets, while reducing assessment costs by over 72% through cascaded evaluation and online weight calibration. PoQ (Proof of Quality) is DGrid's proprietary consensus mechanism designed to prevent low-quality model deployment and data manipulation at the protocol layer.
Disclaimer: The information on this page may come from third-party sources and is for reference only. It does not represent the views or opinions of Gate and does not constitute any financial, investment, or legal advice. Virtual asset trading involves high risk. Please do not rely solely on the information on this page when making decisions. For details, see the Disclaimer.
Comment
0/400
No comments