6 papers
Post-Training, Finetuning, and Steering:A Field Guide for Curating Training Data
Maria Attarian, Yifan Ruan, Claas Voelcker +5
While Vision-Language-Action models have enabled impressive cross-task transfer in zero-shot settings, \emph{adaptation} methods are often necessary to achieve consistent performan…
Push-Wiper: Toward General-Purpose Robotic Cleaning across Varied Stains and Surfaces with Segmented Pushing Trajectories
Renhao Lu, Mingxin Wang, Chenyang Cao +5
Viscous stains, characterized by high viscosity and complex rheological properties, remain a major challenge for robotic surface cleaning. Conventional wiping often spreads the sta…
QPILOTS: Efficient Test-Time Q-Steering for Flow Policies
Yifan Ruan, Chenyang Cao, Andreas Burger +7
Flow-matching and diffusion policies are expressive action generators, but optimizing them with temporal-difference reinforcement learning (RL) remains difficult. Effective policy…
Residual Reward Models: Leveraging Prior Knowledge for Efficient Preference-based Reinforcement Learning in Robotics
Chenyang Cao, Miguel Rogel-García, Mohamed Nabail +2
Preference-based Reinforcement Learning (PbRL) provides a promising alternative to heuristic reward design in complex robotic environments. However, PbRL often suffers from poor sa…
FOSP: Fine-tuning Offline Safe Policy through World Models
Chenyang Cao, Yucheng Xin, Silang Wu +4
Offline Safe Reinforcement Learning (RL) seeks to address safety constraints by learning from static datasets and restricting exploration. However, these approaches heavily rely on…
Offline Goal-Conditioned Reinforcement Learning for Safety-Critical Tasks with Recovery Policy
Chenyang Cao, Zichen Yan, Renhao Lu +2
Offline goal-conditioned reinforcement learning (GCRL) aims at solving goal-reaching tasks with sparse rewards from an offline dataset. While prior work has demonstrated various ap…