Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

VLM-AR3L: Vision-Language Models for Absolute and Relative Rewards in Reinforcement Learning

About

Designing effective reward functions remains a major challenge in reinforcement learning (RL), particularly in open-ended environments where task goals are abstract and difficult to quantify. In this work, we present VLM-AR3L, a framework that leverages Vision-Language Models (VLMs) to provide both absolute and relative rewards for RL. VLM-AR3L interprets an agent's visual observations in the context of a natural language task goal, and learns both absolute and relative rewards from VLM-generated preference labels. The absolute reward model predicts scalar evaluations for individual states, while the relative reward model compares consecutive observations to infer progress or regression toward the task goal. Their integration combines the stability of state-based evaluation with the robustness of comparative supervision. We evaluate VLM-AR3L across benchmarks spanning classic control, manipulation, and open-world embodied tasks, with a particular focus on Minecraft given its visual complexity and long-horizon decision-making requirements. Experimental results show that VLM-AR3L consistently outperforms prior VLM-based reward learning methods.

Kuan-Chen Chen, Winston Chen, Wei-Fang Sun, Min-Chun Hu• 2026

Related benchmarks

TaskDatasetResultRank
Drawer-OpenMetaWorld
Success Rate100
14
sweep-intoMetaWorld
Success Rate100
14
soccerMetaWorld
Success Rate100
14
Shear sheepMineDojo
Success Rate (%)70
12
Cart-Pole ControlCart Pole
Success Rate100
11
Milk CowMineDojo
Success Rate95
6
Combat SpiderMineDojo
Success Rate85
6
Hunt CowMineDojo
Success Rate52
6
Pass WaterMetaWorld
Success Rate100
5
Straighten RopeSoftGym
Success Rate100
5
Showing 10 of 10 rows

Other info

Follow for update