# LARY: A Latent Action Representation Yielding Benchmark for Generalizable Vision-to-Action Alignment

arXiv: https://arxiv.org/abs/2604.11689 · HTML: https://mlsanity.com/paper/2604.11689/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Placed in Human-video and latent-action pretraining, one of 13 directions in Vision-language-action models.
- Builds on 31 of its 35 references that are in the map.
- First posted on arXiv 13 Apr 2026.

## Placed in

- Vision-language-action models / [Human-video and latent-action pretraining](https://mlsanity.com/vla/human-video-and-latent-action-pretraining/)

## Builds on (most cited first, 30 of 31)

- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [Ego4D: Around the World in 3,000 Hours of Egocentric Video](https://mlsanity.com/paper/2110.07058/) (2021)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning](https://mlsanity.com/paper/2306.03310/) (2023)
- [GR00T N1: An Open Foundation Model for Generalist Humanoid Robots](https://mlsanity.com/paper/2503.14734/) (2025)
- [SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features](https://mlsanity.com/paper/2502.14786/) (2025)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)
- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks](https://mlsanity.com/paper/2112.03227/) (2021)
- [UniVLA: Learning to Act Anywhere with Task-centric Latent Actions](https://mlsanity.com/paper/2505.06111/) (2025)
- [AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems](https://mlsanity.com/paper/2503.06669/) (2025)
- [Latent Action Pretraining from Videos](https://mlsanity.com/paper/2410.11758/) (2024)
- [The EPIC-KITCHENS Dataset: Collection, Challenges and Baselines](https://mlsanity.com/paper/2005.00343/) (2020)
- [EgoDex: Learning Dexterous Manipulation from Large-Scale Egocentric Video](https://mlsanity.com/paper/2505.11709/) (2025)
- [VLABench: A Large-Scale Benchmark for Language-Conditioned Robotics Manipulation with Long-Horizon Reasoning Tasks](https://mlsanity.com/paper/2412.18194/) (2024)
- [Open-MAGVIT2: An Open-Source Project Toward Democratizing Auto-regressive Visual Generation](https://mlsanity.com/paper/2409.04410/) (2024)
- [DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos](https://mlsanity.com/paper/2602.06949/) (2026)
- [Learning to Act without Actions](https://mlsanity.com/paper/2312.10812/) (2023)
- [AdaWorld: Learning Adaptable World Models with Latent Actions](https://mlsanity.com/paper/2503.18938/) (2025)
- [TACO: Benchmarking Generalizable Bimanual Tool-ACtion-Object Understanding](https://mlsanity.com/paper/2401.08399/) (2024)
- [RoboCOIN: An Open-Sourced Bimanual Robotic Data Collection for Integrated Manipulation](https://mlsanity.com/paper/2511.17441/) (2025)
- [IGOR: Image-GOal Representations are the Atomic Control Units for Foundation Models in Embodied AI](https://mlsanity.com/paper/2411.00785/) (2024)
- [EWMBench: Evaluating Scene, Motion, and Semantic Quality in Embodied World Models](https://mlsanity.com/paper/2505.09694/) (2025)
- [MGSampler: An Explainable Sampling Strategy for Video Action Recognition](https://mlsanity.com/paper/2104.09952/) (2021)
- [VideoWorld: Exploring Knowledge Learning from Unlabeled Videos](https://mlsanity.com/paper/2501.09781/) (2025)
- [CoMo: Learning Continuous Latent Motion from Internet Videos for Scalable Robot Learning](https://mlsanity.com/paper/2505.17006/) (2025)
- [What Do Latent Action Models Actually Learn?](https://mlsanity.com/paper/2506.15691/) (2025)
- [Latent Action Pretraining Through World Modeling](https://mlsanity.com/paper/2509.18428/) (2025)

## Built on by (most cited first, 12 of 12)

- [From Foundation to Application: Improving VLA Models in Practice](https://mlsanity.com/paper/2607.06403/) (2026)
- [RotVLA: Rotational Latent Action for Vision-Language-Action Model](https://mlsanity.com/paper/2605.13403/) (2026)
- [Is Forward Prediction Enough? Physical State Grounding for JEPA World Models](https://mlsanity.com/paper/2608.06799/) (2026)
- [RynnWorld-4D: 4D Embodied World Models for Robotic Manipulation](https://mlsanity.com/paper/2607.06559/) (2026)
- [Robot Learning from Human Videos: A Survey](https://mlsanity.com/paper/2604.27621/) (2026)
- [What Matters for Latent Actions in Robot Learning](https://mlsanity.com/paper/2608.19613/) (2026)
- [Towards Predictive, Aligned, and Scalable Robot Learning](https://mlsanity.com/paper/2607.11270/) (2026)
- [LAFP: Preserving Latent Action Structure in Latent Policy Learning via Flow Matching](https://mlsanity.com/paper/2606.10517/) (2026)
- [WLA3: World Latent Action Modeling for Semantics, Dynamics, and Kinematics](https://mlsanity.com/paper/2609.15870/) (2026)
- [GeoLAM: Learning Geometry-Grounded Latent Actions from Unlabeled Human Videos](https://mlsanity.com/paper/2609.17099/) (2026)
- [CogWAM: Aligning Semantic Cognition with World Action Modeling via Event-Driven Interfaces](https://mlsanity.com/paper/2609.37721/) (2026)
- [LAPT-Scale: Scaling Embodied Policies with Latent Action Pre-Training on Human Videos](https://mlsanity.com/paper/2610.13732/) (2026)

## Abstract (arXiv)

While the shortage of explicit action data limits Vision-Language-Action (VLA) models, human action videos offer a scalable yet unlabeled data source. A critical challenge in utilizing large-scale human video datasets lies in transforming visual signals into ontology-independent representations, known as latent actions. However, the capacity of latent action representation to derive robust control from visual observations has yet to be rigorously evaluated. We introduce the Latent Action Representation Yielding (LARY) Benchmark, a unified framework for evaluating latent action representations on both high-level semantic actions (what to do) and low-level robotic control (how to do). The comprehensively curated dataset encompasses over one million videos (1,000 hours) spanning 151 action categories, alongside 620K image pairs and 595K motion trajectories across diverse embodiments and environments. Our experiments reveal two crucial insights: (i) General visual foundation models, trained without any action supervision, consistently outperform specialized embodied latent action models. (ii) Latent-based visual space is fundamentally better aligned to physical action space than pixel-based space. These results suggest that general visual representations inherently encode action-relevant knowledge for physical control, and that semantic-level abstraction serves as a fundamentally more effective pathway from vision to action than pixel-level reconstruction.