/TECHStanford researchers release HumanLM, an RL framework that aligns LLM states with human beliefs and emotions to simulate usersJZDY#120The researchers also launched Humanual, a benchmark for user simulation.Combined views992 posts, first seen 10h ago14 repostsStanford researchers release HumanLM, an RL framework that aligns LLM states with human beliefs and emotions to simulate usersJZDY#120The researchers also launched Humanual, a benchmark for user simulation.Combined views992 posts, first seen 10h ago14 repostsUseful Links