arXiv:2611.07679 · v1
PhySA: Decoupling Aesthetic and Physical Rewards for Humanoid Locomotion
摘要
Humanoid locomotion policies trained without an explicit smoothness objective tend to produce jerky, high-energy gaits that stress actuators. Vision-language models (VLMs) offer an appealing reward source: given a rendered rollout, a VLM returns an aesthetic quality score on [0,10] without task-specific feature engineering. We find, however, that aesthetic preference and physical smoothness point in opposite directions: across four production VLMs, clips with higher joint acceleration receive higher aesthetic scores (mean Pearson r=+0.66). We study PhySA (Physical Smoothness and Aesthetics), which keeps the two on separate, independently removable reward channels: a sparse, asynchronous VLM score and a phase-triggered correction channel that adjusts physical penalty weights from simulator state. Ablations attribute the smoothness gain to the correction channel. On MuJoCo Humanoid-v5 at 2.6M steps, PhySA lowers joint acceleration by 26 to 32% relative to vanilla PPO, Text2Reward, and Eureka (three seeds), at the cost of shorter episodes. In a single extended run it lowers joint acceleration by 59.5% relative to Prompt2Policy, while a three-seed replication shows a smaller, non-significant reduction (7.2%). The gain does not transfer to IsaacLab or the Unitree G1, even after rescaling the smoothness term.
作者备注：24 pages, 12 figures, 11 tables. Accepted to the CoRL 2026 Workshop on R2RL: Roadmap to Sample-Efficient Real-World Reinforcement Learning