# Multi-focus Attention Network for Efficient Deep Reinforcement Learning

arXiv: https://arxiv.org/abs/1712.04603 · HTML: https://mlsanity.com/paper/1712.04603/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Dec 2017.

## Built on by (most cited first, 10 of 10)

- [Joint Attention for Multi-Agent Coordination and Social Learning](https://mlsanity.com/paper/2104.07750/) (2021)
- [Goal-Aware Cross-Entropy for Multi-Target Reinforcement Learning](https://mlsanity.com/paper/2110.12985/) (2021)
- [Temporal-Spatial Causal Interpretations for Vision-Based Reinforcement Learning](https://mlsanity.com/paper/2112.03020/) (2021)
- [Leveraging Reward Consistency for Interpretable Feature Discovery in Reinforcement Learning](https://mlsanity.com/paper/2309.01458/) (2023)
- [FlowRL: Flow-Augmented Few-Shot Reinforcement Learning for Semi-Structured Sensor Data](https://mlsanity.com/paper/2409.14178/) (2024)
- [Goal-Conditioned Q-Learning as Knowledge Distillation](https://mlsanity.com/paper/2208.13298/) (2022)
- [Interpretability for Conditional Coordinated Behavior in Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2304.10375/) (2023)
- [Pay Attention to What and Where? Interpretable Feature Extractor in Vision-based Deep Reinforcement Learning](https://mlsanity.com/paper/2504.10071/) (2025)
- [Optimizing Sensory Neurons: Nonlinear Attention Mechanisms for Accelerated Convergence in Permutation-Invariant Neural Networks for Reinforcement Learning](https://mlsanity.com/paper/2506.00691/) (2025)
- [ZeroDVFS: Zero-Shot LLM-Guided Core and Frequency Allocation for Embedded Platforms](https://mlsanity.com/paper/2601.08166/) (2026)

## Abstract (arXiv)

Deep reinforcement learning (DRL) has shown incredible performance in learning various tasks to the human level. However, unlike human perception, current DRL models connect the entire low-level sensory input to the state-action values rather than exploiting the relationship between and among entities that constitute the sensory input. Because of this difference, DRL needs vast amount of experience samples to learn. In this paper, we propose a Multi-focus Attention Network (MANet) which mimics human ability to spatially abstract the low-level sensory input into multiple entities and attend to them simultaneously. The proposed method first divides the low-level input into several segments which we refer to as partial states. After this segmentation, parallel attention layers attend to the partial states relevant to solving the task. Our model estimates state-action values using these attended partial states. In our experiments, MANet attains highest scores with significantly less experience samples. Additionally, the model shows higher performance compared to the Deep Q-network and the single attention model as benchmarks. Furthermore, we extend our model to attentive communication model for performing multi-agent cooperative tasks. In multi-agent cooperative task experiments, our model shows 20% faster learning than existing state-of-the-art model.