# MEMBOT: Memory-Based Robot in Intermittent POMDP

arXiv: https://arxiv.org/abs/2509.11225 · HTML: https://mlsanity.com/paper/2509.11225/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 18 references that are in the map.
- First posted on arXiv 14 Sep 2025.

## Builds on (most cited first, 10 of 10)

- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning](https://mlsanity.com/paper/1910.10897/) (2019)
- [What Matters in Learning from Offline Human Demonstrations for Robot Manipulation](https://mlsanity.com/paper/2108.03298/) (2021)
- [Deep Recurrent Q-Learning for Partially Observable MDPs](https://mlsanity.com/paper/1507.06527/) (2015)
- [Scene Memory Transformer for Embodied Agents in Long-Horizon Tasks](https://mlsanity.com/paper/1903.03878/) (2019)
- [KalMamba: Towards Efficient Probabilistic State Space Models for RL under Uncertainty](https://mlsanity.com/paper/2406.15131/) (2024)
- [Learning Deep Neural Network Policies with Continuous Memory States](https://mlsanity.com/paper/1507.01273/) (2015)
- [Constrained Average-Reward Intermittently Observable MDPs](https://mlsanity.com/paper/2504.13823/) (2025)
- [Optimizing pre-scheduled, intermittently-observed MDPs](https://mlsanity.com/paper/2305.09105/) (2023)

## Built on by (most cited first, 1 of 1)

- [UniMem: Unifying Multimodal Memory and Control for Vision-Language-Action Models](https://mlsanity.com/paper/2608.22869/) (2026)

## Abstract (arXiv)

Robotic systems deployed in real-world environments often operate under conditions of partial and often intermittent observability, where sensor inputs may be noisy, occluded, or entirely unavailable due to failures or environmental constraints. Traditional reinforcement learning (RL) approaches that assume full state observability are ill-equipped for such challenges. In this work, we introduce MEMBOT, a modular memory-based architecture designed to address intermittent partial observability in robotic control tasks. MEMBOT decouples belief inference from policy learning through a two-phase training process: an offline multi-task learning pretraining stage that learns a robust task-agnostic latent belief encoder using a reconstruction losses, followed by fine-tuning of task-specific policies using behavior cloning. The belief encoder, implemented as a state-space model (SSM) and a LSTM, integrates temporal sequences of observations and actions to infer latent state representations that persist even when observations are dropped. We train and evaluate MEMBOT on 10 robotic manipulation benchmark tasks from MetaWorld and Robomimic under varying rates of observation dropout. Results show that MEMBOT consistently outperforms both memoryless and naively recurrent baselines, maintaining up to 80% of peak performance under 50% observation availability. These findings highlight the effectiveness of explicit belief modeling in achieving robust, transferable, and data-efficient policies for real-world partially observable robotic systems.