Vision-Language-Action models have recently emerged as a powerful paradigm for general-purpose robot learning, enabling agents to map visual observations and natural-language instructions into executable robotic actions. Though popular, they are primarily trained via supervised fine-tuning or training-time reinforcement learning, requiring explicit fine-tuning phases, human interventions, or controlled data collection. Consequently, existing methods remain unsuitable for challenging simulated- or physical-world deployments, where robots must respond autonomously and flexibly to evolving environments. To address this limitation, we introduce a Test-Time Reinforcement Learning for VLAs (TT-VLA), a framework that enables on-the-fly policy adaptation during inference. TT-VLA formulates a dense reward mechanism that leverages step-by-step task-progress signals to refine action policies during test time while preserving the SFT/RL-trained priors, making it an effective supplement to current VLA models. Empirical results show that our approach enhances overall adaptability, stability, and task success in dynamic, previously unseen scenarios under simulated and real-world settings. We believe TT-VLA offers a principled step toward self-improving, deployment-ready VLAs.

As text-to-motion (T2M) models move beyond short action captions toward real-world applications, they are increasingly expected to follow diverse instructions. However, existing evaluation protocols remain insufficient for systematically assessing instruction following, physical plausibility, and fine-grained controllability under such diverse language instructions. To address this gap, we introduce T2MBench, a benchmark with 1,025 prompts across four categories—Dynamics, Complexity, Interaction, and Accuracy—covering 14 representative T2M models and three evaluation dimensions: LLM-Based Evaluation, Multi-Factor Motion Evaluation, and Fine-grained Accuracy Evaluation. Our results reveal substantial gaps in current T2M models’ ability to consistently follow diverse language instructions while maintaining motion plausibility and fine-grained control.T2MBench establishes a reusable diagnostic benchmark for advancing research toward more reliable T2M generation under diverse language instructions. Data and materials are available at https://github.com/Courteous121/T2MBench-release.

Multimodal Large Language Models (MLLMs) demonstrate remarkable performance across a wide range of domains, with increasing emphasis on enhancing their zero-shot generalization capabilities for unseen tasks across various modalities. Instruction tuning has emerged as an effective strategy for achieving zero-shot generalization by finetuning pretrained models on diverse multimodal tasks. As the scale of MLLMs continues to grow, parameter-efficient finetuning becomes increasingly critical. However, most existing parameter-efficient approaches focus only on single modalities and often overlook the multimodal characteristics during finetuning. In this work, we introduce a novel Multimodal Prompt Tuning (M²PT) approach for efficient instruction tuning of MLLMs. M²PT effectively integrates visual and textual prompts into the vision encoder and language processor respectively during finetuning, facilitating the extraction and alignment of features across modalities. Empirical results on various multimodal evaluation datasets demonstrate the superior performance of our approach compared to several state-of-the-art baselines. A comprehensive set of ablation studies validates the effectiveness of our prompt design and the efficiency of our approach.