Gamer: A Generalist Vision-Language-Action Foundation Model for Long-Horizon Real-Time Game Agent
arXiv:2611.18397v1 Announce Type: new Abstract: We introduce Gamer, a generalist agent framework built for long-horizon tasks in complex interactive environments. Its end-to-end architecture combines perception, memory management, and multi-step reasoning. A shared training and deployment pipeline supports interaction with both conventional 2D graphical interfaces and complex 3D simulated worlds. We first continue pre-training on a dataset constructed from tens of thousands of hours of device-le
Read original sourceSource record
Project analysis
Project analysis is not yet available. See the source record above.