The paper introduces Gauntlet, a framework that lets large language models autonomously build game-playing agents from a bare contract—just a game description, raw observation/action interface, and an empty policy file. In a single session, the model experiments with the game, compiles a standalone controller, and the resulting program is evaluated on held‑out instances without further model calls. The authors demonstrate that these compiled agents can win full‑scale games such as StarCraft II and Civilization, marking the first time a language‑agent system has achieved standalone victory in such complex titles.
By Joey Xiao, Haonan Huang
arXiv:2610.13995v1 Announce Type: cross
Abstract: Post-training large language models (LLMs) on two-player zero-sum games provides a verifiable reward, and the gains have been shown to transfer to ge...
By Weixian Waylon Li, Jiaxin Zhang, Han Zhou, Chengzu Li, Ke Wang, Fengxiang He, Yiwen Guo
arXiv:2604.07733v2 Announce Type: replace
Abstract: Evaluating strategic decision-making in LLM-based agents requires generative, competitive, and longitudinal environments, yet few benchmarks provid...
By John Chen, Sihan Cheng, Can Gurkan, Mingyi Lin
arXiv:2607. 05904v1 Announce Type: new Abstract: Training a language model against its own reference-free judgments (the premise of self-rewarding, self-play, and LLM-as-a-judge pipelines) assumes a model's verdict on a shown answer tracks correctness.
By Chenyu Zhou
arXiv:2606. 29932v1 Announce Type: new Abstract: Long-horizon strategic planning in complex strategy games demands concurrent reasoning across multiple decision domains under imperfect information and sparse reward.
By Tianyu Jin, Shuo Chen, Yida Wang, Liuyu Xiang, Yingzhuo Liu, Zhiyao Jiang, Yexin Li, Zhaofeng He
arXiv:2607. 06854v1 Announce Type: cross Abstract: Reinforcement learning agents for imperfect-information card games are only as strong as the opponents they train against, and they are hard to grade, since they beat a random opponent over 99 percent of the time and only tie copies of themselves.
By Nima Kelidari, Mohammadsaeed Haghi, Mahdi Salmani
CivBench is an open‑source benchmark that evaluates language‑model agents in the long‑horizon, tool‑mediated game Civilization VI using the Model Context Protocol (MCP). Each episode lasts over 300 turns, generating thousands of tool calls across a 76‑tool action space, and includes a narration layer that translates visual game state into structured text. The study characterises agent behaviour across four model families, introducing Proactive Monitoring Rate (PMR) and RAG@10 as interface‑level metrics, and finds that agents often under‑monitor strategic state and fail to execute near‑term commitments despite tool access and explicit guidance.
By Austin Tudor David Andrews, Liam Wilkinson, Jamie Heagerty, Harry Coppock, Jakob Nicolaus Foerster, Rui Ponte Costa
arXiv:2606. 29932v3 Announce Type: replace Abstract: Long-horizon strategic planning in complex strategy games requires coordinating tightly coupled decision domains, including technology, economy, diplomacy, and military, across hundreds of turns under imperfect information.
By Tianyu Jin, Shuo Chen, Yida Wang, Liuyu Xiang, Yingzhuo Liu, Zhiyao Jiang, Yexin Li, Peipei Li, Zhaofeng He
arXiv:2608. 01425v1 Announce Type: cross Abstract: Training LLM-based multi-agent systems with multi-agent reinforcement learning is rapidly gaining traction, and a parallel line of work argues that such systems should be judged by their behavior, not only their reward.
By Yi Mao, Andrew Perrault
The paper presents a continuous Dyna loop that trains a multi‑agent world model for a ten‑hero MOBA game and evaluates the resulting policy in the real game. By using imagined episodes and real‑game data to continuously refine the model, the policy achieves a 70.2% win rate as radiant on unseen seeds, a dramatic improvement over dream‑only training. The study highlights that internal dream metrics can miss model exploitation, that the world model must be repaired on the policy’s own games, and that the policy’s success stems from macro‑level pressure rather than precise micro‑tactics.
By Jordy Kieto
The paper introduces World Reward Models (WRMs), a new approach for machine‑learning agents that selects the best candidate program directly from execution results instead of predicting intermediate outcomes. WRMs, exemplified by the ForeCritic model, are trained on thousands of executed program pairs and achieve a 79.4% success rate in choosing the better program, outperforming general‑purpose LLM judges. When integrated into systems like AIDE, ML‑Master, and ForeAgent, ForeCritic reduces mean task time by 4.4–7.5× while maintaining or improving overall performance.
By Yuxiao Lee, Zherui Yang, Fan Liu, Hao Liu
The paper investigates the problem of sharing a single critic across multiple parallel environments in reinforcement learning. It shows that when environments assign different expected returns to the same state, a shared critic must reconcile conflicting value targets, which can distort advantage estimates and misguide policy updates. The authors propose a simple fix—providing the critic with the environment index—demonstrating through bandit models and experiments on CartPole, MuJoCo, BipedalWalker, and 16 Procgen games that this conditional critic stabilizes learning and boosts returns, achieving a 40.8% improvement in aggregate normalized return on unseen levels.
By Zhenya Liu, Yang Meng, Zhuokai Zhao, Xuefeng Liu, Yuxin Chen