DiVeR: Decision-Critical Verifier Learning for VLA Test-Time Scaling
Paper • 2610.04933 • Published • 11
EmbodiedAI
-
-
RealtimeWAM: One-Step Asynchronous World Action Models
Paper • 2610.06617 • Published • 16 -
Sparse-WAM: Accelerating World Action Models via Action-Guided Sparse Imagination
Paper • 2609.38984 • Published -
Staircase Policy: Streaming Inference for World-Action Models with Large Action Chunks
Paper • 2609.36471 • Published -
GlanceWAM: Sparse Test-Time Imagination for World-Action Models
Paper • 2608.23927 • Published • 1 -
MotorMind: Scaffolding General Vision Language Models for Zero-Shot Robot Manipulation
Paper • 2609.38078 • Published • 110 -
FlashVLA: Streaming Action Decoding for Fast and Asynchronous VLA Inference
Paper • 2608.27384 • Published -
Long-WAM: Scaling the Context of World-Action Models
Paper • 2610.10528 • Published • 82 -
Streaming-WAM: Action-Conditioned World-Action Model for Asynchronous Robot Manipulation
Paper • 2609.28927 • Published -
Toward Real-Time VLAs: Stage-Aware Two-Step Flow Denoising and System-Level Evaluation
Paper • 2609.39822 • Published • 2 -
Execution-Aligned Progressive Noise for Consistent Asynchronous Replanning in Generative Robot Policies
Paper • 2610.06090 • Published • 2 -
Reactive Real-Time Flow Policies via Asynchronous Distribution Alignment
Paper • 2609.36540 • Published -
RAVEL: Asynchronous Rolling Inference for Flow-Based Vision-Language-Action Models
Paper • 2609.34170 • Published -
StairVLA: Stage-Aware Hierarchical Action Generation for Vision-Language-Action Models
Paper • 2610.07756 • Published