VIMA-Bench

'VIMA-Bench' is a benchmark dataset that contains six tasks used to evaluate the performance of robotic manipulation methods, particularly in the context of self-evolving agents.

Papers using VIMA-Bench (3)

Evolvable Embodied Agent for Robotic Manipulation via Long Short-Term Reflection and Optimization2026Confusion-Aware In-Context-Learning for Vision-Language Models in Robotic Manipulation2026Towards Fast, Memory-based and Data-Efficient Vision-Language Policy2025