SOLAR-RL: Efficient Semi-Online Long-Horizon RL Framework
Discover SOLAR-RL, a novel semi-online reinforcement learning method enhancing long-horizon task completion in GUI agents with efficient training.
QDTraj: Diverse Trajectory Primitives for Robotic Manipulation
Discover QDTraj, a method generating diverse trajectory primitives to boost robotic manipulation of articulated objects in dynamic environments.
ArmSSL: Robust Black-Box Watermarking for SSL Encoders
ArmSSL offers adversarially robust black-box watermarking to secure self-supervised learning encoders with reliable ownership verification.
LLM-Based Grading System for K-12 Non-Native English Learners
Discover an LLM-driven grading system tailored for K-12 non-native English learners, enhancing dialogue generation with adaptive proficiency control.
Nex Playground: Active Gaming Beyond Nintendo & PlayStation
Discover Nex Playground, the bestselling active gaming device that promotes fitness and fun for kids, now at its lowest price of the year.