Research team from Nanjing University proposed FOCUS, a causal model-based offline RL algorithm, which uses causal structure ...

a diffusion-large-language-model-based framework that has been improved through the use of reinforcement learning. The group posted a paper describing their work and features of the new framework ...