Skip to content
Preprint

InternW0-$\Delta$: A World Action Model Bridging Predictive Dynamics and Actions with 20K+ Hours of Open Data

Sep 2026 · 6 citations · ⚡ 2 influential
Computer Science

TL;DR

Causal Imprint is introduced, which learns future-relevant scene changes from training-only future supervision and provides predictive representations directly to the action expert without future-video rollout at inference.

Abstract

World Action Models (WAMs) jointly model visual dynamics and action generation for generalist robot manipulation. A central challenge is to integrate priors from large-scale pretrained models---including visual dynamics, scene semantics, geometry, and motion---into a unified framework for robot action generation. We introduce InternW0-$\Delta$, a unified WAM pretrained on a heterogeneous corpus that outperforms prior methods across simulation benchmarks and real-robot platforms. InternW0-$\Delta$ combines pretrained visual dynamics, scene-level semantics, 4D geometric and motion priors, and action generation within a Mixture-of-Transformers (MoT) framework. A pretrained video expert and an action expert interact under semantic guidance from a frozen VLM, while a pretrained 4D foundation model injects geometric and motion priors through training-only distillation. We further introduce Causal Imprint, which learns future-relevant scene changes from training-only future supervision and provides predictive representations directly to the action expert without future-video rollout at inference. For large-scale joint training, we construct a heterogeneous corpus of robot demonstrations, UMI data, egocentric human demonstrations, and Ego2Robot data, curated and aligned under a common state-action representation. The resulting corpus contains over 20K hours of processed training data, to our knowledge the largest open-source corpus of its kind. We pretrain InternW0-$\Delta$ on this corpus and demonstrate strong performance across simulation benchmarks and real-robot platforms. We will open source the training code, model weights, infrastructure, data-processing pipeline, and processed data where licenses permit. Project page: https://internrobotics.github.io/InternW0-Delta/

View source

Similar papers

Preprint Sep 2026

GE-Act 2.0: Pretraining and Scaling a World-Action Model for Robotic Manipulation

Genie Envisioner Act 2.0 (GE-Act 2.0) is introduced, a world-action model whose trainable generative and action components are all initialized from scratch on manipulation data, and its components are jointly trained with knowledge-aligned selective optimization (KASO), which reduces mismatched supervision.

AgiBot Research Team, Ren-Hang Liu, Wen-Zhi Zhao et al. · 4 citations · ⚡1
Preprint Oct 2026

UniWAM: Unified World-Action Model

UniWAM is introduced, a unified architecture that integrates a physical reasoner, a world generator, and an action predictor to jointly learn semantic understanding of the physical world, visual generation, and action prediction and uncovers a log-linear scaling law of unified human-robot co-training.

Wen-Xuan Song, Jia-Yi Chen, Jing-Bo Wang et al. · 6 citations · ⚡2
Preprint Aug 2026

JEPA-WAM: Learning Vision-Language-Action Policies with Joint-Embedding World Modeling

JEPA-WAM, a latent WAM built in a pretrained V-JEPA space, which couples latent transition prediction with continuous action generation through a shared predictor, predicts a spatially structured joint current-future target that captures task-shared visual temporal structure between current and future observations, whi...

Yi-Han Lin, Jia-Wei He, Shi-Feng Bao et al. · 12 citations · ⚡1
Preprint Aug 2026

LD4WAM: Learning Latent Dynamics from Human Videos for World Action Models

LD4WAM is presented, which pairs a Latent Dynamics Model trained with semantic reconstruction and real motion alignment with a World Dynamics Action Model built as a mixture-of-transformers (MoT), which preserves full future-video generation and uses learnable queries to distill these latent dynamics from generated fut...

Zhen Shen, Jia-Qi Liang, Jasper Lu et al. · 3 citations
Preprint Sep 2026

DeltaWAM: Delta World Action Models for Bimanual Manipulation

This work proposes DeltaWAM, which jointly predicts visual deltas and actions using dense-anchor, sparse-delta, and action streams, with three architectures that differ in representation and computation sharing, and develops Streaming Delta Memory (SDM), which updates cached anchor context with compact observed deltas,...

Han Yan, Zi-Shang Xiang, Hao-Kai Jiang et al. · 0 citations
Sep 2026

From World Models to World Action Models: Rethinking Next-State Prediction

CF-WAM is proposed, a dynamic next-state prediction framework that samples visual, semantic, geometric, and interaction projections of the same future, standardizes them into a common video form, and supervises a unified WAM across these projections.

Ting-Yu Yuan, Zi-Ming Ji, Biao-Liang Guan et al. · 1 citation

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.