Beyond Benchmarking: A New Paradigm for Evaluation and Assessment of Large Language Models

Jin Liu, Qingquan Li, Wenlong Du

In current benchmarks for evaluating large language models (LLMs), there are issues such as evaluation content restriction, untimely updates, and lack of optimization guidance. In…