#safety fine-tuning

Benchmarking Argumentative Behaviour of LLMs: A Study ofNEW⭐9

A study benchmarking how large language models use and defend against character attacks in political debate, comparing their argumentative strategies with human...