Live intelligence

Latest AI updates

5 updates

Research

Cycle WAM: Coupling Action and Video through Cycle Consistency

Cycle WAM introduces a unified training framework that treats video and action as paired modalities linked by causality, using bidirectional conversion and a differentiable cycle-reconstruction objective to enforce consistency between forward and inverse dynamics. It achieves state-of-the-art results on LIBERO-Pro and improves DreamZero on DROID, with sharper visual detail in forward-dynamics predictions compared to UWM.

arxiv1d agoScore 78

World Action ModelsVideo ModelingRobot Learning

Research Editor selected

FloMo: 3D Scene Flow as a World Action Model Intermediate for Learning from Human Video

FloMo is a 5B video-generation model fine-tuned to predict 3D scene flow and action tokens, using scene flow as an intermediate representation to transfer video pretraining to robot action prediction. Co-trained on human video and robot teleoperation data, it generalizes to unseen objects and tasks, outperforming video prediction objectives. Experiments show that predicting video wastes model capacity, as action tokens attend more to motion than video.

Why it matters: This paper presents a concrete, recent AI development with empirical evidence that 3D scene flow improves sample efficiency and generalization in robot learning from video.

arxiv1d agoScore 85

Robot learningVideo pretrainingScene flow

Research Editor selected

PyRUA-DataFarm: LLM Agent as Cost-Efficient Teleoperator for Robot Data Collection

PyRUA-DataFarm uses an LLM agent to automatically rebuild robot tasks in simulation, generating real-ready training data without teleoperation. Policies trained on this farm data match or exceed those trained on teleoperated episodes across five real-robot tasks, with lower cost and better robustness to visual changes. The system includes automated certification of expert controllers and rendered episodes.

Why it matters: This presents a concrete, recent AI development with measurable cost and performance gains, offering practical value for robot learning and data collection.

arxiv1d agoScore 85

robot learningdata generationLLM agents

Research

CausalWAM: Learning the Causes of Actions in World Action Models

The paper introduces CausalWAM, a method to ensure that world action models (WAMs) generate actions based only on action-relevant latent variables, not nuisance variables. It proves that under a sparsity constraint, the latent variables used for actions are an invertible function of the true action-relevant latents, identifying the structure up to permutation. The method adds a sparse routing path to a pretrained WAM's action head without changing the world modeling objective. Experiments on synthetic data and real robots show improved closed-loop success.

arxiv3d agoScore 78

world modelscausal representation learningrobot learning

Opinion

MIT Technology Review Examines the Gap Between Humanoid Robot Hype and Real AI Progress

MIT Technology Review reports that while Elon Musk and others predict humanoid robots will soon transform labor and daily life, many robotics researchers remain skeptical. The piece details current systems like Google DeepMind's Gemini Robotics running on ALOHA 2 and Physical Intelligence's π0 series, which can perform some general tasks but fail outside their training data. It contrasts flashy, often teleoperated demos with lab progress toward vision-language-action models and world models. Res

MIT Technology Review AIOct 8, 2026Score 68

humanoid roboticsvision-language-action modelsworld models