Research map / Vision-language and multimodal / Human and embodied modeling
MM-ACT: Learn from Multimodal Parallel Generation to Act
Haotian Liang, Xinyi Chen, Bin Wang, Mingkang Chen, Yitian Liu, Yuhao Zhang, Zanxin Chen, Tianshuo Yang, Yilun Chen, Jiangmiao Pang, Dong Liu, Xiaokang Yang and 3 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add subtask-planning text co-training to action generation (ablation) | MM-ACT · LIBERO-Long | success rate (%): 88.0 -> 93.0 | Improved |
| add text co-training (ablation) | MM-ACT · RoboTwin2.0 (8 tasks) | success rate (%): 43.13% -> 46.5% | Improved |
| add future-image co-training (ablation) | MM-ACT · RoboTwin2.0 (8 tasks) | success rate (%): 43.13% -> 48.75% | Improved |
| add text and image co-training together (ablation) | MM-ACT · RoboTwin2.0 (8 tasks) | success rate (%): 43.13% -> 52.38% | Improved |
| add text co-training (ablation) | MM-ACT · RoboTwin2.0 Lift Pot | success rate (%): 40% -> 28% | Hurt |
| add future-image co-training (ablation) | MM-ACT · RoboTwin2.0 Dump Bin Bigbin | success rate (%): 13% -> 8% | Hurt |
| add text and image co-training (ablation) | MM-ACT · RoboTwin2.0 Place Burger Fries | success rate (%): 46% -> 73% | Improved |
| use MM-ACT instead of OpenVLA-OFT | Franka real-world tasks · Franka real-world (3 tasks) | success rate (%): 58.6 -> 72.0 | Improved |
| use MM-ACT instead of π0 | Franka real-world tasks · Franka real-world (3 tasks) | success rate (%): 70.0 -> 72.0 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DreamVLA — DreamVLA: A Vision-Language-Action Model Dreamed with Comprehensive World Knowledge (NeurIPS 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Mastering Robot Manipulation with Multimodal Prompts through Pretraining and Multi-task Fine-tuning (ICML 2024 · similarity 0.89)
- FineVLA: Fine-Grained Instruction Alignment for Steerable Vision-Language-Action Policies (NeurIPS 2026 · similarity 0.89)
- ATI-VLA: Action-Centric Predictive Vision–Language–Action Models via Actionable Alignment Then Adaptive Injection (NeurIPS 2026 · similarity 0.88)
- M*: A Modular, Extensible, Serving System for Multimodal Models (NeurIPS 2026 · similarity 0.88)
- ActionUNet: Improving Robustness of VLA Models with Efficient Multi-scale Fine-tuning (NeurIPS 2026 · similarity 0.88)
- ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
