Remember Smarter: Visual History Compressor and Hyperbolic Experience Space for Robotic Memory
On the map: Vision-language-action models → Reasoning and chain-of-thought in VLA
Abstract (arXiv)
Long-horizon robot policies require compact access to recent observations and reusable experience without expanding the vision-language-action (VLA) context. We introduce Remember Smarter (RS), a plug-and-play module with complementary visual-history and hyperbolic experience-memory branches. Its visual branch compresses multi-view patch histories using bidirectional spatial Mamba and causal temporal Mamba, then exposes the resulting memory to action-facing hidden states through residual cross-attention while leaving the VLM visual-token stream unchanged. Its experience branch stores successful final-layer VLM states in a Poincare VAE space, organizes them hierarchically, and asynchronously converts retrieved experience into geodesic prompt tokens without blocking action inference. When adapted to pi0, RS increases total success on LIBERO-Plus from 53.6% to 70.6% and achieves substantial performance gains in real-robot experiments designed to evaluate memory retention and experience utilization.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Reasoning and chain-of-thought in VLA, one of 13 directions in Vision-language-action models. Builds on 15 of its 33 references that are in the map. First posted on arXiv 15 Aug 2026.
Builds on
15 of 33 references.
Closest ancestors · 2
Same direction, newest first
- 2025Rethinking Progression of Memory State in Robotic Manipulation: An Object-Centric Perspective14 Nov
- 2025MAP-VLA: Memory-Augmented Prompting for Vision-Language-Action Model in Robotic Manipulation12 Nov
Other references · 13
- 2017Poincar'e Embeddings for Learning Hierarchical Representationscited by 684
- 2018Learning Continuous Hierarchies in the Lorentz Model of Hyperbolic Geometrycited by 265
- 2018Hyperbolic Entailment Cones for Learning Hierarchical Embeddingscited by 162
- 2025VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Modelcited by 145
- 2019Continuous Hierarchical Representations with Poincar'e Variational Auto-Encoderscited by 112
- 2024RoboMamba: Efficient Vision-Language-Action Model for Robotic Reasoning and Manipulationcited by 103
- 2020From Trees to Continuous Embeddings and Back: Hyperbolic Hierarchical Clusteringcited by 65
- 2020Differentiating through the Fr'echet Meancited by 53
- 2025STORM: Token-Efficient Long Video Understanding for Multimodal LLMscited by 49
- 2024Model Compression Method for S4 with Diagonal State Space Layers using Balanced Truncationcited by 7
- 2024Episodic Memory Verbalization using Hierarchical Representations of Life-Long Robot Experiencecited by 4
- 2025State-Space Hierarchical Compression with Gated Attention and Learnable Sampling for Hour-Long Video Understanding in Large Multimodal Modelscited by 3
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies
- 2026T2Mem: Learning Test-Time Memory for Robotics ↗
- 2025MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulation
- 2026MemoryVLA++: Temporal Modeling via Memory and Imagination in Vision-Language-Action Models
- 2026Memory Without Curation: Learning Multi-Stream Latent Memory for Robot Policies
- 2025MemER: Scaling Up Memory for Robot Control via Experience Retrieval
- 2026NativeMEM: Native Memory Compression for Long-Horizon Robotic Manipulation
- 2026Remember what you did?: Learning Behavioral Memories for Partially Observable Object Manipulation