The Year in AI Papers
The year/Independent research

Paper 2602.12099

GigaBrain-0.5M*: a VLA That Learns From World Model-Based Reinforcement Learning

GigaBrain Team, Boyuan Wang, Bohan Li, Chaojun Ni, Guan Huang, Guosheng Zhao, Hao Li, Jie Li, Jindi Lv, Jingyu Liu, Lv Feng, Mingming Yu, et al.

Published
Feb 2026
Research lab
Independent
Citations
7
GitHub
Not linked

01 In brief

Summary

GigaBrain-0.5M* is a vision-language-action (VLA) model trained via world model-based reinforcement learning, built on GigaBrain-0.5, which is pretrained on over 10,000 hours of robotic manipulation data and ranks first on the RoboChallenge benchmark.

The model integrates RAMP (Reinforcement leArning via world Model-conditioned Policy), a four-stage iterative pipeline: world model pretraining, policy training with world model conditioning, human-in-the-loop rollout data collection, and continual training.

RAMP conditions the policy on future state latents and value estimates predicted by a world model, providing richer information than the sparse advantage signals used in RECAP, which is shown to be a special case of RAMP.

Experiments show GigaBrain-0.5 achieves state-of-the-art success rates on eight internal tasks and 30 RoboChallenge tasks, with an intermediate version scoring 51.67% average success.

RAMP outperforms baselines like AWR and RECAP, yielding approximately 30% improvements on tasks such as Laundry Folding, Box Packing, and Espresso Preparation, and enables reliable long-horizon execution in real-world deployments.

02 From the paper

Abstract

Vision-language-action (VLA) models that directly predict multi-step action chunks from current observations face inherent limitations due to constrained scene understanding and weak future anticipation capabilities. In contrast, video world models pre-trained on web-scale video corpora exhibit robust spatiotemporal reasoning and accurate future prediction, making them a natural foundation for enhancing VLA learning. Therefore, we propose \textit{GigaBrain-0.5M*}, a VLA model trained via world model-based reinforcement learning. Built upon \textit{GigaBrain-0.5}, which is pre-trained on over 10,000 hours of robotic manipulation data, whose intermediate version currently ranks first on the international RoboChallenge benchmark. \textit{GigaBrain-0.5M*} further integrates world model-based reinforcement learning via \textit{RAMP} (Reinforcement leArning via world Model-conditioned Policy) to enable robust cross-task adaptation. Empirical results demonstrate that \textit{RAMP} achieves substantial performance gains over the RECAP baseline, yielding improvements of approximately 30\% on challenging tasks including \texttt{Laundry Folding}, \texttt{Box Packing}, and \texttt{Espresso Preparation}. Critically, \textit{GigaBrain-0.5M$^*$} exhibits reliable long-horizon execution, consistently accomplishing complex manipulation tasks without failure as validated by real-world deployment videos on our \href{https://gigabrain05m.github.io}{project page}.