behavior

Generative AI

Liquid AI’s LFM2-2.6B-Exp uses pure reinforcement learning RL and dynamic hybrid reasoning to optimize small model behavior

Generative AI

OpenAI has trained its LLMs to accept bad behavior