One paper on “Benchmarking Legal Knowledge of Large Language Models” is on arxiv, welcome to LawBench for more details.