# Attentional Bottleneck: Towards an Interpretable Deep Driving Network

arXiv: https://arxiv.org/abs/2005.04298 · HTML: https://mlsanity.com/paper/2005.04298/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 May 2020.

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

- [Explainability of deep vision-based autonomous driving systems: Review and challenges](https://mlsanity.com/paper/2101.05307/) (2021)
- [A Survey on Interpretable Reinforcement Learning](https://mlsanity.com/paper/2112.13112/) (2021)
- [Exploring Imitation Learning for Autonomous Driving with Feedback Synthesizer and Differentiable Rasterization](https://mlsanity.com/paper/2103.01882/) (2021)
- [Goal-Guided Transformer-Enabled Reinforcement Learning for Efficient Autonomous Navigation](https://mlsanity.com/paper/2301.00362/) (2023)
- [Safety Implications of Explainable Artificial Intelligence in End-to-End Autonomous Driving](https://mlsanity.com/paper/2403.12176/) (2024)
- [Beyond Imitation: A Life-long Policy Learning Framework for Path Tracking Control of Autonomous Driving](https://mlsanity.com/paper/2404.17198/) (2024)
- [Interpretable Decision-Making for End-to-End Autonomous Driving](https://mlsanity.com/paper/2508.18898/) (2025)
- [Scene-Aware Explainable Multimodal Trajectory Prediction](https://mlsanity.com/paper/2410.16795/) (2024)
- [Pay Attention to What and Where? Interpretable Feature Extractor in Vision-based Deep Reinforcement Learning](https://mlsanity.com/paper/2504.10071/) (2025)
- [Driving the Wrong Way: Leveraging Interpretability in End2End Autonomous Driving Models](https://mlsanity.com/paper/2607.06328/) (2026)

## Abstract (arXiv)

Deep neural networks are a key component of behavior prediction and motion generation for self-driving cars. One of their main drawbacks is a lack of transparency: they should provide easy to interpret rationales for what triggers certain behaviors. We propose an architecture called Attentional Bottleneck with the goal of improving transparency. Our key idea is to combine visual attention, which identifies what aspects of the input the model is using, with an information bottleneck that enables the model to only use aspects of the input which are important. This not only provides sparse and interpretable attention maps (e.g. focusing only on specific vehicles in the scene), but it adds this transparency at no cost to model accuracy. In fact, we find slight improvements in accuracy when applying Attentional Bottleneck to the ChauffeurNet model, whereas we find that the accuracy deteriorates with a traditional visual attention model.