Detecting Rhetorical Questions in LLMs: A Linear Probing Study
Explore how large language models represent rhetorical questions using linear probing across social media datasets, revealing distinct linguistic cues.
HiVLA: Advanced Visual-Grounded Robotic Manipulation System
Discover HiVLA, a hierarchical system enhancing robotic manipulation with visual grounding and separate planning and control for superior performance.
UI-Zoomer: Adaptive Uncertainty-Driven Zoom for GUI Grounding
Boost GUI element localization with UI-Zoomer, an adaptive zoom-in method using uncertainty quantification. No extra training needed, improves accuracy up...
UMI-3D: Advanced 3D Spatial Perception for Robotics
Discover UMI-3D, enhancing Universal Manipulation Interface with 3D spatial perception using LiDAR for robust robotic data collection.
Token Importance in On-Policy Distillation Explained
Discover how token importance boosts on-policy distillation efficiency by focusing on student entropy and teacher-student divergence signals.