Some Notes on Adversarial Attacks on LLMs (by milosgajdos) — discussion

#ai