[
    {
        "id": "osp-17924",
        "type": "article-journal",
        "title": "VLA-ACL: Action-Consistent Visual Token Pruning for Efficient Vision-Language-Action Models",
        "author": [
            {
                "family": "Du",
                "given": "Owen"
            },
            {
                "family": "Yue",
                "given": "Yang"
            },
            {
                "family": "Zhang",
                "given": "Jie"
            },
            {
                "family": "Pi",
                "given": "Jiaqi"
            },
            {
                "family": "Chen",
                "given": "Chi Bene"
            },
            {
                "family": "Huang",
                "given": "Gao"
            }
        ],
        "URL": "https://omanscience.com/ar/articles/vla-acl-action-consistent-visual-token-pruning-for-efficient-vision-language-action-models",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "Vision-Language-Action (VLA) models achieve strong robotic manipulation performance but incur high computational costs from processing long token sequences at every control step, limiting real-time deployment. Visual token pruning offers a direct solution, as visual patches dominate the input sequence and contain considerable redundancy. Existing approaches, however, either rely on indirect training-free heuristics, such as attention scores and motion thresholds, or require costly fine-tuning of the base VLA model. We introduce VLA-ACL (Action Consistency Learning), which learns a lightweight visual token pruning policy through action-level supervision while keeping the base VLA model entirely frozen. The training objective encourages actions produced from pruned visual contexts to remain consistent with the full-context teacher, with ground-truth actions as auxiliary supervision. This directly ties token selection to its effect on the downstream control output. Experiments on LIBERO and real-world manipulation tasks show that VLA-ACL prunes up to 87.5% of visual tokens while retaining competitive performance, reduces computation by up to 75%, and achieves a 1.5x inference speedup. These results establish a stronger performance-efficiency trade-off than existing frozen-VLA pruning methods and demonstrate the value of action-level supervision for visual token selection. Code is available at https://github.com/du-owen/VLA-ACL."
    }
]