Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement Learning

• URL: http://arxiv.org/abs/2505.03792v2
• Date: Tue, 03 Jun 2025 12:14:42 GMT
• Title: Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement Learning
• Authors: Lang Feng, Weihao Tan, Zhiyi Lyu, Longtao Zheng, Haiyang Xu, Ming Yan, Fei Huang, Bo An,
• Abstract summary: We propose a novel online fine-tuning method, Counterfactual Soft Reinforcement Learning (CoSo), better suited to the textual output space of VLM agents.<n>CoSo prioritizes the exploration of action-critical tokens while reducing the impact of semantically redundant or low-impact tokens.<n>Our results across a diverse set of agent tasks, including Android device control, card gaming, and embodied AI, highlight its remarkable ability to enhance exploration efficiency and deliver consistent performance gains.
• Score: 45.05706458955535
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Online fine-tuning vision-language model (VLM) agents with reinforcement learning (RL) has shown promise for equipping agents with multi-step, goal-oriented capabilities in dynamic environments. However, their open-ended textual action space and non-end-to-end nature of action generation present significant challenges to effective online exploration in RL, e.g., explosion of the exploration space. We propose a novel online fine-tuning method, Counterfactual Soft Reinforcement Learning (CoSo), better suited to the textual output space of VLM agents. Compared to prior methods that assign uniform uncertainty to all tokens, CoSo leverages counterfactual reasoning to dynamically assess the causal influence of individual tokens on post-processed actions. By prioritizing the exploration of action-critical tokens while reducing the impact of semantically redundant or low-impact tokens, CoSo enables a more targeted and efficient online rollout process. We provide theoretical analysis proving CoSo's convergence and policy improvement guarantees, and extensive empirical evaluations supporting CoSo's effectiveness. Our results across a diverse set of agent tasks, including Android device control, card gaming, and embodied AI, highlight its remarkable ability to enhance exploration efficiency and deliver consistent performance gains. The code is available at https://github.com/langfengQ/CoSo.

Related papers

• Application of LLM Guided Reinforcement Learning in Formation Control with Collision Avoidance [1.1718316049475228]
  Multi-Agent Systems (MAS) excel at accomplishing complex objectives through the collaborative efforts of individual agents.<n>In this paper, we introduce a novel framework that aims to overcome the challenge of designing an effective reward function.<n>By giving large language models (LLMs) on the prioritization of tasks, our framework generates reward functions that can be dynamically adjusted online.
  arXiv  Detail & Related papers  (2025-07-22T09:26:00Z)
• RALLY: Role-Adaptive LLM-Driven Yoked Navigation for Agentic UAV Swarms [15.891423894740045]
  We develop a Role-Adaptive LLM-Driven Yoked navigation algorithm RALLY.<n>RALLY uses structured natural language for efficient semantic communication and collaborative reasoning.<n> Experiments show that RALLY outperforms conventional approaches in terms of task coverage, convergence speed, and generalization.
  arXiv  Detail & Related papers  (2025-07-02T05:44:17Z)
• Scaling Autonomous Agents via Automatic Reward Modeling And Planning [52.39395405893965]
  Large language models (LLMs) have demonstrated remarkable capabilities across a range of tasks.<n>However, they still struggle with problems requiring multi-step decision-making and environmental feedback.<n>We propose a framework that can automatically learn a reward model from the environment without human annotations.
  arXiv  Detail & Related papers  (2025-02-17T18:49:25Z)
• Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System [75.25394449773052]
  Large Language Model (LLM) based multi-agent systems (MAS) show remarkable potential in collaborative problem-solving.<n>Yet they still face critical challenges: low communication efficiency, poor scalability, and a lack of effective parameter-updating optimization methods.<n>We present Optima, a novel framework that addresses these issues by significantly enhancing both communication efficiency and task effectiveness.
  arXiv  Detail & Related papers  (2024-10-10T17:00:06Z)
• Compromising Embodied Agents with Contextual Backdoor Attacks [69.71630408822767]
  Large language models (LLMs) have transformed the development of embodied intelligence. This paper uncovers a significant backdoor security threat within this process. By poisoning just a few contextual demonstrations, attackers can covertly compromise the contextual environment of a black-box LLM.
  arXiv  Detail & Related papers  (2024-08-06T01:20:12Z)
• Reinforcing Language Agents via Policy Optimization with Action Decomposition [36.984163245259936]
  This paper proposes decomposing language agent optimization from the action level to the token level. We then derive the Bellman backup with Action Decomposition (BAD) to integrate credit assignments for both intra-action and inter-action tokens. Implementing BAD within the PPO algorithm, we introduce Policy Optimization with Action Decomposition (POAD)
  arXiv  Detail & Related papers  (2024-05-23T14:01:44Z)
• Fine-Tuning Large Vision-Language Models as Decision-Making Agents via Reinforcement Learning [79.38140606606126]
  We propose an algorithmic framework that fine-tunes vision-language models (VLMs) with reinforcement learning (RL) Our framework provides a task description and then prompts the VLM to generate chain-of-thought (CoT) reasoning. We demonstrate that our proposed framework enhances the decision-making capabilities of VLM agents across various tasks.
  arXiv  Detail & Related papers  (2024-05-16T17:50:19Z)
• Efficient Reinforcement Learning via Decoupling Exploration and Utilization [6.305976803910899]
  Reinforcement Learning (RL) has achieved remarkable success across multiple fields and applications, including gaming, robotics, and autonomous vehicles. In this work, our aim is to train agent with efficient learning by decoupling exploration and utilization, so that agent can escaping the conundrum of suboptimal Solutions. The above idea is implemented in the proposed OPARL (Optimistic and Pessimistic Actor Reinforcement Learning) algorithm.
  arXiv  Detail & Related papers  (2023-12-26T09:03:23Z)
• PsiPhi-Learning: Reinforcement Learning with Demonstrations using Successor Features and Inverse Temporal Difference Learning [102.36450942613091]
  We propose an inverse reinforcement learning algorithm, called emphinverse temporal difference learning (ITD) We show how to seamlessly integrate ITD with learning from online environment interactions, arriving at a novel algorithm for reinforcement learning with demonstrations, called $Psi Phi$-learning.
  arXiv  Detail & Related papers  (2021-02-24T21:12:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.