# Latent Spatial Memory for Video World Models

Page: https://stenobird.com/podcast/daily-paper-cast-7079649/latent-spatial-memory-for-video-world-models
Text version: https://stenobird.com/podcast/daily-paper-cast-7079649/latent-spatial-memory-for-video-world-models.md
Podcast: [Daily Paper Cast](https://stenobird.com/podcast/daily-paper-cast-7079649)
Published: 2026-06-10T04:34:53+00:00
Episode link: https://share.transistor.fm/s/54ecc822
Audio file: https://media.transistor.fm/54ecc822/a9e0d13f.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/latent-spatial-memory-for-video-world-models
Duration seconds: 1504

## Resource

🤗 Upvotes: 49 | cs.CV Authors: Weijie Wang, Haoyu Zhao, Yifan Yang, Feng Chen, Zeyu Zhang, Yefei He, Zicheng Duan, Donny Y. Chen, Yuqing Yang, Bohan Zhuang Title: Latent Spatial Memory for Video World Models Arxiv: http://arxiv.org/abs/2606.09828v1 Abstract: Video world models that maintain 3D spatial consistency across generated frames typically rely on explicit point cloud memory constructed in RGB space. This design is both computationally expensive, requiring repeated rendering and VAE encoding, and inherently lossy, as the round trip through pixel space discards rich features of the learned latent representation. In this paper, we introduce \emph{latent spatial memory} for video world models, a persistent 3D cache that stores scene information directly in the diffusion latent space, avoiding pixel-space reconstruction. Building on this, we propose Mirage, a latent-space spatial memory framework that constructs the memory by lifting latent tokens into 3D via depth-guided back-projection and queries it by synthesizing novel views through direct latent-space warping. This unified formulation eliminates both the information loss of pixel-space reconstruction and the computational burden of repeated encoding and rendering. Experiments show that latent spatial memory achieves up to \textbf{10.57}$\times$ faster end-to-end video generation and \textbf{55}$\times$ reduction in memory footprint relative to explicit 3D baselines. Leveraging the geometric prior of the diffusion model, Mirage attains state-of-the-art performance on WorldScore and strong reconstruction quality on RealEstate10K.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/latent-spatial-memory-for-video-world-models/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/daily-paper-cast-7079649/latent-spatial-memory-for-video-world-models.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.