Toward Comprehensive Benchmarking of the Biological Knowledge of Frontier Large Language Models比较生物知识基准,并考察其评估前沿模型滥用风险的适用性。核对原文署名 ↗