Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text Paper • 2607.21072 • Published 1 day ago • 34
VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon Paper • 2607.01804 • Published 23 days ago • 31
GFT: From Imitation to Reward Fine-Tuning with Unbiased Group Advantages and Dynamic Coefficient Rectification Paper • 2604.14258 • Published Apr 15 • 23
ClawGUI: A Unified Framework for Training, Evaluating, and Deploying GUI Agents Paper • 2604.11784 • Published Apr 13 • 143