Reconstructing Is Not Acting: Action-Centric Latent Dynamics Modeling

• URL: http://arxiv.org/abs/2609.15189v1
• Date: Mon, 14 Sep 2026 08:10:46 GMT
• Title: Reconstructing Is Not Acting: Action-Centric Latent Dynamics Modeling
• Abstract summary: Latent action models learn action representations from unlabeled videos by inferring latent actions from visual transitions and reconstructing future states.<n>We propose $textbfACT-LAM, a lightweight action-centric framework that strengthens both action extraction and action utilization.<n>Experiments on several robotic datasets and the VP$2$ benchmark demonstrate stronger latent action consistency, forward dynamics, and downstream visual planning performance.
• Score: 21.941208569455657
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Latent action models (LAMs) learn action representations from unlabeled videos by inferring latent actions from visual transitions and reconstructing future states. However, we identify a fundamental $\textbf{reconstruction-action mismatch}$: lower reconstruction error does not necessarily yield better latent dynamics or downstream performance. We attribute this mismatch to two underconstrained aspects of reconstruction-based latent dynamics modeling: (i) the inverse dynamics model (IDM) is not explicitly encouraged to distinguish action-related transitions from nuisance appearance, and (ii) the forward dynamics model (FDM) can underutilize the inferred latent action by exploiting predictive shortcuts from the current state. To address both limitations, we propose $\textbf{ACT-LAM}$, a lightweight action-centric framework that strengthens both action extraction and action utilization. Specifically, its Action Query IDM (AQ-IDM) employs learnable action queries and gated aggregation to selectively extract rich action-related transition cues without strong information bottlenecks. And its Action Token FDM (AT-FDM) projects latent actions into action tokens that progressively interact with evolving state representations, enabling continuous state-aware action conditioning. ACT-LAM further streamlines feature processing to concentrate model capacity on latent dynamics modeling. Extensive experiments on several robotic datasets and the VP$^2$ benchmark demonstrate stronger latent action consistency, forward dynamics, and downstream visual planning performance with fewer trainable parameters and lower computational overhead. In particular, ACT-LAM surpasses the previous state of the art by $\textbf{7.6%}$ on the aggregated VP$^2$ success rate. Codes at $\href{https://github.com/DingjieFu/ACT-LAM}{url}$.

Related papers

• Overcoming Statistical Bias in Action-Controllable World Models [18.890337378174678]
  Action-conditioned world models aim to predict how visual environments evolve under an agent's actions.<n>Future frames are often highly predictable from visual inertia and recurring motion patterns alone.<n>We introduce CoCo, a Counterfactual Consistency framework to enhance action controllability.
  arXiv  Detail & Related papers  (2026-08-05T10:10:36Z)
• Causally Debiased Latent Action Model for Embodied Action Conditioned World Models [47.281442841934876]
  Action-conditioned world models (ACWMs) aim to simulate future observations conditioned on embodied actions.<n>Existing LAMs are typically trained with reconstruction-only objectives.<n>We propose CD-LAM, a causally debiased framework for LAM-based ACWMs.
  arXiv  Detail & Related papers  (2026-07-10T08:20:27Z)
• SCAR: Self-Supervised Continuous Action Representation Learning [36.917304453471864]
  We propose SCAR, a joint inverse-forward dynamics framework for learning unified action representations across embodiments from visual transitions.<n>Built on a pretrained generative backbone, SCAR uses an inverse dynamics model (IDM) to infer latent actions from latent observation pairs and a forward dynamics model (FDM) to predict future dynamics conditioned on them.<n>Experiments on the Procgen and Robotwin dataset show that the learned unified latent action representation serves as a stronger conditioning interface for world modeling than embodiment-specific raw actions.
  arXiv  Detail & Related papers  (2026-05-13T16:23:11Z)
• RotVLA: Rotational Latent Action for Vision-Language-Action Model [54.22746299071677]
  We introduce RotVLA, a VLA framework built on a continuous rotational latent action representation.<n>Latent actions are modeled as elements of SO(n), providing continuity, compositionality, and structured geometry aligned with real-world action dynamics.<n>RotVLA consists of a VLM backbone and a flow-matching action head, pretrained on large-scale cross-embodiment robotic datasets and human videos.
  arXiv  Detail & Related papers  (2026-05-13T11:58:02Z)
• From Imagined Futures to Executable Actions: Mixture of Latent Actions for Robot Manipulation [88.39072412680633]
  We propose MoLA, a control-oriented interface that transforms imagined future videos into executable representations.<n>We evaluate our approach on simulated benchmarks and real-world robot manipulation tasks.
  arXiv  Detail & Related papers  (2026-05-12T14:15:16Z)
• InstrAct: Towards Action-Centric Understanding in Instructional Videos [12.356484522873577]
  InstrAction is a pretraining framework for instructional videos' action-centric representations.<n>We first introduce a data-driven strategy, which filters noisy captions and generates action-centric hard negatives.<n>At the visual feature level, an Action Perceiver extracts motion-relevant tokens from redundant video encodings.
  arXiv  Detail & Related papers  (2026-04-09T20:51:13Z)
• SPIRAL: A Closed-Loop Framework for Self-Improving Action World Models via Reflective Planning Agents [135.00390535239129]
  We introduce SPIRAL, a self-improving planning and iterative reflective action world modeling framework.<n> SPIRAL formulates ActWM as a closed-loop think-act-reflect process, where generation proceeds step by step under explicit planning and feedback.<n> Experiments across multiple TI2V backbones demonstrate consistent gains on ActWM-Bench and mainstream video generation benchmarks.
  arXiv  Detail & Related papers  (2026-03-09T14:00:36Z)
• MOTIF: Learning Action Motifs for Few-shot Cross-Embodiment Transfer [55.982504915794514]
  Cross-embodiment policies typically rely on shared-private architectures.<n>We introduce MOTIF for efficient few-shot cross-embodiment transfer.<n>We show that MOTIF significantly outperforms strong baselines in few-shot transfer scenarios.
  arXiv  Detail & Related papers  (2026-02-14T13:21:40Z)
• \ extsc{NaVIDA}: Vision-Language Navigation with Inverse Dynamics Augmentation [50.027425808733994]
  textscNaVIDA is a unified VLN framework that couples policy learning with action-grounded visual dynamics and adaptive execution.<n>textscNaVIDA augments training with chunk-based inverse-dynamics supervision to learn causal relationship between visual changes and corresponding actions.<n>Experiments show that textscNaVIDA achieves superior navigation performance compared to state-of-the-art methods with fewer parameters.
  arXiv  Detail & Related papers  (2026-01-26T06:16:17Z)
• DyMoDreamer: World Modeling with Dynamic Modulation [52.27044216359359]
  A critical bottleneck in deep reinforcement learning (DRL) is sample inefficiency, as training high-performance agents often demands extensive environmental interactions.<n>We introduce DyMoDreamer, a novel algorithm that incorporates a dynamic modulation mechanism to improve the extraction of dynamic features and enrich the temporal information.<n>Experiments demonstrate that DyMoDreamer sets a new state-of-the-art on the Atari $100$k benchmark with a $156.6$% mean human-normalized score.
  arXiv  Detail & Related papers  (2025-09-29T13:54:42Z)
• Capturing Rich Behavior Representations: A Dynamic Action Semantic-Aware Graph Transformer for Video Captioning [13.411096520754507]
  Existing video captioning methods merely provide shallow or simplistic representations of object behaviors.<n>We propose a dynamic action semantic-aware graph transformer to comprehensively capture the essence of object behavior.
  arXiv  Detail & Related papers  (2025-02-19T14:16:47Z)

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.