Think in Games: Learning to Reason in Games via Reinforcement Learning with Large Language Models Article Swipe

MAKE IMPACT

Liao Yi , Yu Gu , Yuan Sui , Zining Zhu , Yifan Lu , Guohua Tang , Zhongqian Sun , Wei Yang · YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2508.21365

Large language models (LLMs) excel at complex reasoning tasks such as mathematics and coding, yet they frequently struggle with simple interactive tasks that young children perform effortlessly. This discrepancy highlights a critical gap between declarative knowledge (knowing about something) and procedural knowledge (knowing how to do something). Although traditional reinforcement learning (RL) agents can acquire procedural knowledge through environmental interaction, they often operate as black boxes and require substantial training data. In contrast, LLMs possess extensive world knowledge and reasoning capabilities, but are unable to effectively convert this static knowledge into dynamic decision-making in interactive settings. To address this challenge, we propose Think in Games (TiG), a novel framework that empowers LLMs to develop procedural understanding through direct interaction with game environments, while retaining their inherent reasoning and explanatory abilities. Specifically, TiG reformulates RL-based decision-making as a language modeling task: LLMs generate language-guided policies, which are refined iteratively through online reinforcement learning based on environmental feedback. Our experimental results show that TiG successfully bridges the gap between declarative and procedural knowledge, achieving competitive performance with dramatically lower data and computational demands compared to conventional RL methods. Moreover, TiG provides step-by-step natural language explanations for its decisions, greatly improving transparency and interpretability in complex interactive tasks.

Type

preprint

Language

en

Landing Page

http://arxiv.org/abs/2508.21365

PDF

https://arxiv.org/pdf/2508.21365

OA Status

green

OpenAlex ID

https://openalex.org/W4415989343

Raw OpenAlex JSON

OpenAlex ID

https://openalex.org/W4415989343

Canonical identifier for this work in OpenAlex

DOI

https://doi.org/10.48550/arxiv.2508.21365

Digital Object Identifier

Title

Think in Games: Learning to Reason in Games via Reinforcement Learning with Large Language Models

Work title

Type

preprint

OpenAlex work type

Language

en

Primary language

Publication year

2025

Year of publication

Publication date

2025-08-29

Full publication date if available

Authors

Liao Yi, Yu Gu, Yuan Sui, Zining Zhu, Yifan Lu, Guohua Tang, Zhongqian Sun, Wei Yang

List of authors in order

Landing page

https://arxiv.org/abs/2508.21365

Publisher landing page

PDF URL

https://arxiv.org/pdf/2508.21365

Direct link to full text PDF

Open access

Yes

Whether a free full text is available

OA status

green

Open access status per OpenAlex

OA URL

https://arxiv.org/pdf/2508.21365

Direct OA link when available

Cited by

0

Total citation count in OpenAlex

Full payload