7 papers
Flex-: A Multi-Stream World-Action Model with Compute Flexibility
Ge Yan, Jinghao Liu, Yuzhi Fan +4
World-action models (WAMs) predict the future to act better, but nearly all of them predict only RGB latents, trained purely for pixel reconstruction, with no explicit signal for t…
Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisons
Anthony Liang, Yigit Korkmaz, Jiahui Zhang +14
General-purpose robot reward models are typically trained to predict absolute task progress from expert demonstrations, providing only local, frame-level supervision. While effecti…
TMRL: Diffusion Timestep-Modulated Pretraining Enables Exploration for Efficient Policy Finetuning
Matthew M. Hong, Jesse Zhang, Anusha Nagabandi +1
Fine-tuning pre-trained robot policies with reinforcement learning (RL) often inherits the bottlenecks introduced by pre-training with behavioral cloning (BC), which produces narro…
HAND Me the Data: Fast Robot Adaptation via Hand Path Retrieval
Matthew Hong, Anthony Liang, Kevin Kim +4
We hand the community HAND, a simple and time-efficient method for teaching robots new manipulation tasks through human hand demonstrations. Instead of relying on task-specific rob…
PEEK: Guiding and Minimal Image Representations for Zero-Shot Generalization of Robot Manipulation Policies
Jesse Zhang, Marius Memmel, Kevin Kim +6
Robotic manipulation policies often fail to generalize because they must simultaneously learn where to attend, what actions to take, and how to execute them. We argue that high-lev…
ReWiND: Language-Guided Rewards Teach Robot Policies without New Demonstrations
Jiahui Zhang, Yusen Luo, Abrar Anwar +5
We introduce ReWiND, a framework for learning robot manipulation tasks solely from language instructions without per-task demonstrations. Standard reinforcement learning (RL) and i…