Research map / Reinforcement learning / Visual and representation RL
Generative Temporal Models with Spatial Memory for Partially Observed Environments
Marco Fraccaro, Danilo Rezende, Yori Zwols, Alexander Pritzel, S. M. Ali Eslami, Fabio Viola
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use a low-dimensional state-space model for agent dynamics and a DND to store position–visual-feature pairs | GTM-SM · CelebA image navigation | long-term prediction coherence | Improved |
| Use a rotation-aware state transition and periodic angle distances for memory retrieval | GTM-SM · Labyrinth rotating agent | prediction coherence | Improved |
| Model momentum, friction, and displacement saturation in the state transition | GTM-SM · Labyrinth walking agent | state prediction accuracy | Improved |
| Add a regression loss using true displacements as targets | GTM-SM walking-agent model · Labyrinth walking agent | transition learning | Improved |
| Use retrieved views in a Generative Query Network prior to combine views across projective transformations | GTM-SM · Labyrinth rotating agent | prediction coherence for unseen viewpoints | Improved |
| Test the single-room-trained model in larger multi-room environments | GTM-SM walking-agent model · Labyrinth multi-room environments | memory of previously visited room textures | Improved |
| Generate corridors between rooms using a model trained only on single rooms | GTM-SM walking-agent model · Labyrinth multi-room environments | corridor rendering | Hurt |
| Generate complex objects with the VAE in the low-resolution Labyrinth images | GTM-SM walking-agent model · Labyrinth walking agent | object rendering | Hurt |
Closest papers by meaning
- Learning Latent Dynamic Robust Representations for World Models (ICML 2024 · similarity 0.89)
- Imagine Beyond ! Distributionally Robust Autoencoding for State Space Coverage in Online Reinforcement Learning (NeurIPS 2025 · similarity 0.88)
- Learning 3D Persistent Embodied World Models (NeurIPS 2025 · similarity 0.88)
- Zero-shot World Models via Search in Memory (NeurIPS 2025 · similarity 0.87)
- Memo: Training Memory-Efficient Embodied Agents with Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Open-World Reinforcement Learning over Long Short-Term Imagination (ICLR 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
