Modality-Autoregressive World-Action Models

• URL: http://arxiv.org/abs/2609.17524v1
• Date: Tue, 15 Sep 2026 17:56:24 GMT
• Title: Modality-Autoregressive World-Action Models
• Abstract summary: We introduce ModAR, the first WAM to autoregressively denoise multiple future modalities before predicting actions.<n>We train from scratch to systematically study how training-data mixtures, predicted modalities, and WAM formulations affect performance.
• Score: 44.96048693516864
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: World-action models (WAMs) jointly model future observations and actions, typically predicting the future as RGB images. Other visual modalities such as depth, pretrained visual features, and point tracks can more efficiently capture geometric, semantic, and motion features. However, how best to combine these modalities within WAMs remains an open question. We introduce ModAR, the first WAM to autoregressively denoise multiple future modalities before predicting actions. This allows each prediction to condition on previously generated modalities. We train from scratch to systematically study how training-data mixtures, predicted modalities, and WAM formulations affect performance. In our evaluations, WAMs benefit from predicting point tracks, DINO features, and depth maps, while additionally predicting future RGB does not provide a consistent benefit. We also find that ModAR's sequential generation outperforms existing WAM formulations, with the highest average success rate at all evaluated data scales. We also fine-tune the video-model-initialized WAM Flex-$π$ on the same data; ModAR achieves a slightly higher observed average success rate (75% vs. 72%) while using approximately $20\times$ fewer training FLOPs and no pretraining. On three real-world bimanual tasks, ModAR outperforms baselines and improves with human videos.

Related papers

• Spatially Aware World Action Model via Geometric Latent Diffusion [42.13843953705695]
  World Action Models (WAMs) leverage the capabilities of large-scale pretrained video diffusion models to jointly predict future observations and actions.<n>We introduce a Spatially Aware World Action Model (SA-WAM) which repurposes a pretrained video model for joint action, RGB, and depth prediction.<n>SA-WAM achieves state-of-the-art results on the RoboCasa and LIBERO-Plus benchmarks, while simultaneously improving future-state predictions.
  arXiv  Detail & Related papers  (2026-09-02T12:42:37Z)
• Disentangled Robot Learning via Separate Forward and Inverse Dynamics Pretraining [28.30092786035367]
  DeFI is a novel framework that Decouples visual Forward and Inverse dynamics pretraining to exploit respective data sources.<n>We introduce the General Forward Dynamics Model (GFDM), pretrained on diverse human and robot videos for future prediction, and the General Inverse Dynamics Model (GIDM), trained via self-supervised learning to infer latent actions from unlabeled video transitions.<n>Experiments on CALVIN ABC-D and SimplerEnv demonstrate state-of-the-art performance, with DeFI achieving an average task length of 4.51 for CALVIN, 51.2% success rate on SimplerEnv-Frac
  arXiv  Detail & Related papers  (2026-03-27T17:20:10Z)
• VFMF: World Modeling by Forecasting Vision Foundation Model Features [67.09340259579761]
  We introduce a generative forecaster that performs autoregressive flow matching in vision foundation models feature space.<n>We show that this latent information more effectively than previously used PCA-based alternatives, both for forecasting and other applications.<n>With matched architecture and compute, our method produces sharper and more accurate predictions than regression across all modalities.
  arXiv  Detail & Related papers  (2025-12-12T02:10:05Z)
• Intention-Conditioned Flow Occupancy Models [80.42634994902858]
  Large-scale pre-training has fundamentally changed how machine learning research is done today.<n>Applying this same framework to reinforcement learning is appealing because it offers compelling avenues for addressing core challenges in RL.<n>Recent advances in generative AI have provided new tools for modeling highly complex distributions.
  arXiv  Detail & Related papers  (2025-06-10T15:27:46Z)
• Enhancing Training Data Attribution with Representational Optimization [57.61977909113113]
  Training data attribution methods aim to measure how training data impacts a model's predictions.<n>We propose AirRep, a representation-based approach that closes this gap by learning task-specific and model-aligned representations explicitly for TDA.<n>AirRep introduces two key innovations: a trainable encoder tuned for attribution quality, and an attention-based pooling mechanism that enables accurate estimation of group-wise influence.
  arXiv  Detail & Related papers  (2025-05-24T05:17:53Z)
• MM-RLHF: The Next Step Forward in Multimodal LLM Alignment [59.536850459059856]
  We introduce MM-RLHF, a dataset containing $mathbf120k$ fine-grained, human-annotated preference comparison pairs.<n>We propose several key innovations to improve the quality of reward models and the efficiency of alignment algorithms.<n>Our approach is rigorously evaluated across $mathbf10$ distinct dimensions and $mathbf27$ benchmarks.
  arXiv  Detail & Related papers  (2025-02-14T18:59:51Z)
• EDA: Evolving and Distinct Anchors for Multimodal Motion Prediction [27.480524917596565]
  We introduce a novel paradigm, named Evolving and Distinct Anchors (EDA), to define the positive and negative components for multimodal motion prediction based on mixture models. EDA enables anchors to evolve and redistribute themselves under specific scenes for an enlarged regression capacity.
  arXiv  Detail & Related papers  (2023-12-15T02:55:24Z)
• Back2Future: Leveraging Backfill Dynamics for Improving Real-time Predictions in Future [73.03458424369657]
  In real-time forecasting in public health, data collection is a non-trivial and demanding task. 'Backfill' phenomenon and its effect on model performance has been barely studied in the prior literature. We formulate a novel problem and neural framework Back2Future that aims to refine a given model's predictions in real-time.
  arXiv  Detail & Related papers  (2021-06-08T14:48:20Z)

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.