VLABench
PulseAugur coverage of VLABench — every cluster mentioning VLABench across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New framework RynnVLA-Latent scales robotic manipulation models using latent actions
Researchers have developed RynnVLA-Latent, a new framework designed to scale the pretraining of vision-language-action (VLA) models for robotic manipulation. This approach utilizes latent actions derived from diverse vi…
-
New VLA framework uses "imagination" for efficient robotic control
Researchers have developed IG-VLA, a novel framework for Vision-Language-Action (VLA) models that enhances robotic manipulation by enabling models to "imagine" future scene evolutions. This approach, detailed in a recen…
-
GaussMemory: New 3D Gaussian Splatting for Task-Driven Robotic Memory
Researchers have introduced GaussMemory, a novel approach to robotic manipulation that utilizes 3D Gaussian Splatting for task-driven spatial memory. Unlike previous passive systems, GaussMemory actively learns which ob…
-
New DIRECT framework optimizes compute for embodied AI planners
Researchers have developed a new framework called DIRECT to optimize the allocation of computational resources for embodied AI planners. The system analyzes multimodal scene context to intelligently route compute, impro…
-
New framework detects robot execution failures using trajectory data
Researchers have developed a new framework called Hide-and-Seek to improve the reliability of robots using Vision-Language-Action (VLA) models. This method detects execution failures by identifying specific actions that…