Scaling Automatic Research Agents via World Models

Researchers propose World Model RL (WMRL), a method that replaces environment execution with a world model to overcome the bottleneck of scaling reinforcement learning for automatic research agents. WMRL accelerates training by 3-4x across various tasks and agent scales, while exceeding standard RL baselines, and enables 4B and 9B agents to outperform much larger open-weight agents of 48B and 120B on held-out benchmarks.

arXiv:2608.12564v1 Announce Type: new Abstract: Automating empirical research is a long-standing direction of AI. Recent automatic research AutoResearch agents bring this goal within reach, as modern LLMs show the capability to independently implement solutions and learn from the execution outcomes. Behind these gains, post-training especially RL plays a central role. In this paper, we identify a fundamental tension when scaling RL for these agents: the two components of every AutoResearch trajectory agent generation and environment execution scale in very different manners, since all generation shares compute through batching, while each execution occupies its exclusive sandbox and real machine time. As a result, the environment execution dominates the training cost and becomes the bottleneck as trajectories grow. To resolve this tension, we propose World Model RL WMRL , which replaces environment execution with a world model to remove this bottleneck. Additionally, the world model can be imperfect, as its rewards are corrupted by bias and noise. Therefore, we further equip WMRL with two mitigations, Online Debiasing and Inverse-Variance Denoising, which offset the bias and suppress the noise respectively. Theoretically, we prove that both mitigations of WMRL strictly improve the convergence guarantee. Empirically, WMRL accelerates training by 3-4x on various tasks at different agent scales, while exceeding the performance of standard RL baselines. Moreover, our post-trained 4B and 9B agents outperform much larger open-weight agents of 48B and 120B on held-out benchmarks. Beyond AutoResearch, WMRL also transfers to post-training embodied VLA policies, which demonstrates the generalizability of our method.