Xirui Jiang, Yuxi Chen, Rui Yang, Feng Luo, Pengcheng Wang, Bingxuan Li, Zihao Zhao, Yushan Li (University of Illinois at Urbana-Champaign) · Gameplay Experience Scaling

Game2Player

Game2Player is a framework in which a Builder agent adapts web games, exposes their internal state and defines verifiable tasks, while a Coding Player and a VLM Player attempt them and feed problems back for repair. The authors build 568 verified web games with 3,436 tasks, collect 15.7K trajectories (307 hours of gameplay) and hold out 50 games as the PlayerBench benchmark. Training the Game2Player-9B VLM agent with progress-weighted SFT raises task progress from 14.3% to 31.2% on PlayerBench and from 30.0% to 53.1% on the external GameWorld benchmark.

Open the source ↗

Similar Research & Safety releases

Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.