A new method, RISED, is introduced for training LLM agents across diverse interactive environments, focusing on prompt-group selection and handling varying learning rates between environments.
An LLM judge tags each rollout using a predefined rubric vocabulary shared across environments.
Across model backbones, RISED achieves the highest mean pass rate across environments and ranks first or second in every individual environment.
Abstract: Training a single LLM agent jointly across diverse interactive environments has attracted increasing attention as a route to generalist agents. Existing curriculum and data-selection strategies often allocate training at the environment level or prioritize local reward-based signals, without explicitly considering relationships between current rollouts across environments for prompt-gro…
