Introducing Neuromodulation in Deep Neural Networks to Learn Adaptive Behaviours
This work addresses the challenge of making intelligent machines more adaptable to unpredictable environments, which is an incremental step in improving artificial systems for tasks like navigation.
The paper tackled the problem of enabling deep neural networks to learn adaptive behaviors by drawing inspiration from biological neuromodulation, and demonstrated that their neuromodulation-based architecture improves adaptation in navigation benchmarks within a meta-reinforcement learning context, showing competitive performance compared to state-of-the-art methods.
Animals excel at adapting their intentions, attention, and actions to the environment, making them remarkably efficient at interacting with a rich, unpredictable and ever-changing external world, a property that intelligent machines currently lack. Such an adaptation property relies heavily on cellular neuromodulation, the biological mechanism that dynamically controls intrinsic properties of neurons and their response to external stimuli in a context-dependent manner. In this paper, we take inspiration from cellular neuromodulation to construct a new deep neural network architecture that is specifically designed to learn adaptive behaviours. The network adaptation capabilities are tested on navigation benchmarks in a meta-reinforcement learning context and compared with state-of-the-art approaches. Results show that neuromodulation is capable of adapting an agent to different tasks and that neuromodulation-based approaches provide a promising way of improving adaptation of artificial systems.