# VAGEN: Reinforcing World Model Reasoning for Multi-Turn VLM Agents

arXiv: https://arxiv.org/abs/2510.16907 · HTML: https://mlsanity.com/paper/2510.16907/

## Key facts

- Cited by 62 papers in the map (data through 14 Oct 2026).
- Placed in World models for language agents, one of 11 directions in World models.
- Builds on 47 of its 82 references that are in the map.
- First posted on arXiv 19 Oct 2025.

## Placed in

- World models / [World models for language agents](https://mlsanity.com/wm/world-models-for-language-agents/)
- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 30 of 47)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [Fine-Tuning Language Models from Human Preferences](https://mlsanity.com/paper/1909.08593/) (2019)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020)
- [Gemini: A Family of Highly Capable Multimodal Models](https://mlsanity.com/paper/2312.11805/) (2023)
- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [Aligning Large Multimodal Models with Factually Augmented RLHF](https://mlsanity.com/paper/2309.14525/) (2023)
- [SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training](https://mlsanity.com/paper/2501.17161/) (2025)
- [VLM-R1: A Stable and Generalizable R1-style Large Vision-Language Model](https://mlsanity.com/paper/2504.07615/) (2025)
- [Group-in-Group Policy Optimization for LLM Agent Training](https://mlsanity.com/paper/2505.10978/) (2025)
- [Training Language Models to Self-Correct via Reinforcement Learning](https://mlsanity.com/paper/2409.12917/) (2024)
- [Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents](https://mlsanity.com/paper/2302.01560/) (2023)
- [RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement Learning](https://mlsanity.com/paper/2504.20073/) (2025)
- [R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization](https://mlsanity.com/paper/2503.10615/) (2025)
- [DreamSim: Learning New Dimensions of Human Visual Similarity using Synthetic Data](https://mlsanity.com/paper/2306.09344/) (2023)
- [Mobile-Agent: Autonomous Multi-Modal Mobile Device Agent with Visual Perception](https://mlsanity.com/paper/2401.16158/) (2024)
- [AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning](https://mlsanity.com/paper/2505.24298/) (2025)
- [ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RL](https://mlsanity.com/paper/2402.19446/) (2024)
- [EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents](https://mlsanity.com/paper/2502.09560/) (2025)

## Built on by (most cited first, 30 of 62)

- [Agent0-VL: Exploring Self-Evolving Agent for Tool-Integrated Vision-Language Reasoning](https://mlsanity.com/paper/2511.19900/) (2025)
- [Qwen-AgentWorld: Language World Models for General Agents](https://mlsanity.com/paper/2606.24597/) (2026)
- [ProRL Agent: Rollout-as-a-Service for RL Training of Multi-Turn LLM Agents](https://mlsanity.com/paper/2603.18815/) (2026)
- [RAGEN-2: Reasoning Collapse in Agentic RL](https://mlsanity.com/paper/2604.06268/) (2026)
- [Visual Generation Unlocks Human-Like Reasoning through Multimodal World Models](https://mlsanity.com/paper/2601.19834/) (2026)
- [Scaling Agentic Reinforcement Learning for Tool-Integrated Reasoning in VLMs](https://mlsanity.com/paper/2511.19773/) (2025)
- [Current Agents Fail to Leverage World Model as Tool for Foresight](https://mlsanity.com/paper/2601.03905/) (2026)
- [GUI-Libra: Training Native GUI Agents to Reason and Act with Action-aware Supervision and Partially Verifiable RL](https://mlsanity.com/paper/2602.22190/) (2026)
- [ENACT: Evaluating Embodied Cognition with World Modeling of Egocentric Interaction](https://mlsanity.com/paper/2511.20937/) (2025)
- [PROGRESSLM: Towards Progress Reasoning in Vision-Language Models](https://mlsanity.com/paper/2601.15224/) (2026)
- [Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond](https://mlsanity.com/paper/2604.22748/) (2026)
- [VisGym: Diverse, Customizable, Scalable Environments for Multimodal Agents](https://mlsanity.com/paper/2601.16973/) (2026)
- [Imagine-then-Plan: Agent Learning from Adaptive Lookahead with World Models](https://mlsanity.com/paper/2601.08955/) (2026)
- [Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application](https://mlsanity.com/paper/2606.12191/) (2026)
- [Beyond Next-Observation Prediction: Agent-Authored World Modeling for Sequential Decision Making](https://mlsanity.com/paper/2606.25421/) (2026)
- [Abstract 3D Perception for Spatial Intelligence in Vision-Language Models](https://mlsanity.com/paper/2511.10946/) (2025)
- [AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents](https://mlsanity.com/paper/2605.17933/) (2026)
- [Generalization in Online Reinforcement Learning for Mobile Agents](https://mlsanity.com/paper/2603.07432/) (2026)
- [PhyWorld: Physics-Faithful World Model for Video Generation](https://mlsanity.com/paper/2605.19242/) (2026)
- [OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents](https://mlsanity.com/paper/2606.02031/) (2026)
- [CoMAP: Co-Evolving World Models and Agent Policies for LLM Agents](https://mlsanity.com/paper/2606.02372/) (2026)
- [Self-Improving World Modelling with Latent Actions](https://mlsanity.com/paper/2602.06130/) (2026)
- [UI-Oceanus: Scaling GUI Agents with Synthetic Environmental Dynamics](https://mlsanity.com/paper/2604.02345/) (2026)
- [Nano World Models: A Minimalist Implementation of Future Video Prediction](https://mlsanity.com/paper/2605.23993/) (2026)
- [Hera: Learning Long-Horizon Coordination for Device-Cloud Collaborative LLM Agents](https://mlsanity.com/paper/2605.24598/) (2026)
- [EnvRL: Learn from Environment Dynamics in Agentic Reinforcement Learning](https://mlsanity.com/paper/2606.17680/) (2026)
- [MEDVISTAGYM: A Scalable Training Environment for Thinking with Medical Images via Tool-Integrated Reinforcement Learning](https://mlsanity.com/paper/2601.07107/) (2026)
- [Recursive Belief Vision Language Action Models](https://mlsanity.com/paper/2602.20659/) (2026)
- [Implicit Turn-Wise Policy Optimization for Proactive User-LLM Interaction](https://mlsanity.com/paper/2603.23550/) (2026)
- [GTR-Turbo: Merged Checkpoint is Secretly a Free Teacher for Agentic VLM Training](https://mlsanity.com/paper/2512.13043/) (2025)

## Abstract (arXiv)

A key challenge in training Vision-Language Model (VLM) agents, compared to Language Model (LLM) agents, lies in the shift from textual states to complex visual observations. This transition introduces partial observability and demands robust world modeling. We ask: Can VLM agents construct internal world models through explicit visual state reasoning? To address this question, we architecturally enforce and reward the agent's reasoning process via reinforcement learning (RL), formulating it as a Partially Observable Markov Decision Process (POMDP). We find that decomposing the agent's reasoning into State Estimation ("what is the current state?") and Transition Modeling ("what comes next?") is critical for success, as demonstrated through five reasoning strategies. Our investigation into how agents represent internal beliefs reveals that the optimal representation is task-dependent: Natural Language excels at capturing semantic relationships in general tasks, while Structured formats are indispensable for precise manipulation and control. Building on these insights, we design a World Modeling Reward that provides dense, turn-level supervision for accurate state prediction, and introduce Bi-Level General Advantage Estimation (Bi-Level GAE) for turn-aware credit assignment. Through this form of visual state reasoning, a 3B-parameter model achieves a score of 0.82 across five diverse agent benchmarks, representing a 3× improvement over its untrained counterpart (0.21) and outperforming proprietary reasoning models such as GPT-5 (0.75), Gemini 2.5 Pro (0.67) and Claude 4.5 (0.62). All experiments are conducted within our VAGEN framework, a scalable system for training and analyzing multi-turn VLM agents in diverse visual environments. Code and data are publicly available at https://vagen-ai.github.io.