PRISM: Preference Refinement via Implicit Scene Modeling for 3D Vision-Language Preference-Based Reinforcement Learning
This addresses the problem of occlusion and short-sighted feedback in robotic environments for scalable, human-aligned reinforcement learning, representing a novel method for a known bottleneck.
The paper tackles the limitations of 2D-based Preference-Based Reinforcement Learning by proposing PRISM, which integrates 3D point cloud modeling and future-aware preference refinement, resulting in significant gains in preference agreement rates and faster policy convergence.
We propose PRISM, a novel framework designed to overcome the limitations of 2D-based Preference-Based Reinforcement Learning (PBRL) by unifying 3D point cloud modeling and future-aware preference refinement. At its core, PRISM adopts a 3D Point Cloud-Language Model (3D-PC-LLM) to mitigate occlusion and viewpoint biases, ensuring more stable and spatially consistent preference signals. Additionally, PRISM leverages Chain-of-Thought (CoT) reasoning to incorporate long-horizon considerations, thereby preventing the short-sighted feedback often seen in static preference comparisons. In contrast to conventional PBRL techniques, this integration of 3D perception and future-oriented reasoning leads to significant gains in preference agreement rates, faster policy convergence, and robust generalization across unseen robotic environments. Our empirical results, spanning tasks such as robotic manipulation and autonomous navigation, highlight PRISM's potential for real-world applications where precise spatial understanding and reliable long-term decision-making are critical. By bridging 3D geometric awareness with CoT-driven preference modeling, PRISM establishes a comprehensive foundation for scalable, human-aligned reinforcement learning.