Echo-Memory: A Controlled Study of Memory in Action World Models
Abstract
Controlled study of memory mechanisms in action-conditioned world models reveals that memory structure and capacity significantly impact open-domain return performance beyond simple replay fidelity measures.
We present Echo-Memory, a controlled study of memory mechanisms in action-conditioned world models. These models generate multi-segment videos from a first frame, text prompt, and camera-action sequence, but their central failure is often memory rather than local image synthesis: after the camera leaves and returns, the scene or salient object may silently change. Existing memory designs are hard to compare because gains are entangled with backbone, training, retrieval, and evaluation differences. Echo-Memory fixes the action-to-video interface and varies only how history is stored and read by the generator. Under a shared video diffusion backbone, optimizer, camera-action representation, sampler, and evaluation pipeline, we compare raw context, compression-based memory, spatial summaries with different read-out paths, and state-space recurrence. This matched matrix separates four otherwise conflated axes: capacity, compression, read-out, and recurrence. We also evaluate memory through a three-branch protocol: replay quality, in-domain loop revisit, and open-domain return probes. The branches routinely disagree, showing that replay fidelity is not a sufficient proxy for remembering a world. Three findings follow. Raw context is a strong capacity baseline and improves open-domain return far more than it improves replay metrics. Compactness is not a free substitute for capacity: aggressive spatial and hybrid-compression memories lose the salient evidence needed for return. Finally, block-wise state-space recurrence is the strongest open-domain return mechanism in our matrix, showing that the structure of implicit memory matters as much as the decision to use it. These results provide a compact protocol for studying memory in action world models beyond isolated replay metrics.
Community
Echo-Memory is a controlled study that systematically evaluates memory mechanisms in action-conditioned world models to address challenges in long-term scene consistency and object retention.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Geometry-Aware Implicit Memory for Video World Models (2026)
- WorldKV: Efficient World Memory with World Retrieval and Compression (2026)
- Robust Dreamer: Deviation-Aware Latent Gaussian Memory for Action-Controlled AR Video Generation (2026)
- Teaching Video Generators to Remember: Eliciting Dynamic Memory for Out-of-Sight State Evolution (2026)
- CausalCine: Real-Time Autoregressive Generation for Multi-Shot Video Narratives (2026)
- LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generation (2026)
- Tensor Memory: Fixed-Size Recurrent State for Long-Horizon Transformers (2026)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Get this paper in your agent:
hf papers read 2606.09803 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 1
Spaces citing this paper 0
No Space linking this paper