# Deep Reinforcement Learning with Modulated Hebbian plus Q Network Architecture

arXiv: https://arxiv.org/abs/1909.09902 · HTML: https://mlsanity.com/paper/1909.09902/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Sep 2019.

## Built on by (most cited first, 5 of 5)

- [A Domain-Agnostic Approach for Characterization of Lifelong Learning Systems](https://mlsanity.com/paper/2301.07799/) (2023)
- [Context Meta-Reinforcement Learning via Neuromodulation](https://mlsanity.com/paper/2111.00134/) (2021)
- [Real-Time Recurrent Reinforcement Learning](https://mlsanity.com/paper/2311.04830/) (2023)
- [The configurable tree graph (CT-graph): measurable problems in partially observable and distal reward environments for lifelong reinforcement learning](https://mlsanity.com/paper/2302.10887/) (2023)
- [Reinforcement Learning With Sparse-Executing Actions via Sparsity Regularization](https://mlsanity.com/paper/2105.08666/) (2021)

## Abstract (arXiv)

This paper presents a new neural architecture that combines a modulated Hebbian network (MOHN) with DQN, which we call modulated Hebbian plus Q network architecture (MOHQA). The hypothesis is that such a combination allows MOHQA to solve difficult partially observable Markov decision process (POMDP) problems which impair temporal difference (TD)-based RL algorithms such as DQN, as the TD error cannot be easily derived from observations. The key idea is to use a Hebbian network with bio-inspired neural traces in order to bridge temporal delays between actions and rewards when confounding observations and sparse rewards result in inaccurate TD errors. In MOHQA, DQN learns low level features and control, while the MOHN contributes to the high-level decisions by associating rewards with past states and actions. Thus the proposed architecture combines two modules with significantly different learning algorithms, a Hebbian associative network and a classical DQN pipeline, exploiting the advantages of both. Simulations on a set of POMDPs and on the MALMO environment show that the proposed algorithm improved DQN's results and even outperformed control tests with A2C, QRDQN+LSTM and REINFORCE algorithms on some POMDPs with confounding stimuli and sparse rewards.