# Deep Abstract Q-Networks

arXiv: https://arxiv.org/abs/1710.00459 · HTML: https://mlsanity.com/paper/1710.00459/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Oct 2017.

## Built on by (most cited first, 6 of 6)

- [Learning Options via Compression](https://mlsanity.com/paper/2212.04590/) (2022)
- [Composing Reinforcement Learning Policies, with Formal Guarantees](https://mlsanity.com/paper/2402.13785/) (2024)
- [Discrete State-Action Abstraction via the Successor Representation](https://mlsanity.com/paper/2206.03467/) (2022)
- [Deep Inductive Logic Programming meets Reinforcement Learning](https://mlsanity.com/paper/2308.16210/) (2023)
- [SAGE: Generating Symbolic Goals for Myopic Models in Deep Reinforcement Learning](https://mlsanity.com/paper/2203.05079/) (2022)
- [Diagnosing and exploiting the computational demands of videos games for deep reinforcement learning](https://mlsanity.com/paper/2309.13181/) (2023)

## Abstract (arXiv)

We examine the problem of learning and planning on high-dimensional domains with long horizons and sparse rewards. Recent approaches have shown great successes in many Atari 2600 domains. However, domains with long horizons and sparse rewards, such as Montezuma's Revenge and Venture, remain challenging for existing methods. Methods using abstraction (Dietterich 2000; Sutton, Precup, and Singh 1999) have shown to be useful in tackling long-horizon problems. We combine recent techniques of deep reinforcement learning with existing model-based approaches using an expert-provided state abstraction. We construct toy domains that elucidate the problem of long horizons, sparse rewards and high-dimensional inputs, and show that our algorithm significantly outperforms previous methods on these domains. Our abstraction-based approach outperforms Deep Q-Networks (Mnih et al. 2015) on Montezuma's Revenge and Venture, and exhibits backtracking behavior that is absent from previous methods.