# Spiking Decision Transformers: Local Plasticity, Phase-Coding, and Dendritic Routing for Low-Power Sequence Control

## TL;DR

This paper introduces SNN-DT, a spiking neural network model for sequence control that achieves comparable performance to traditional transformers while drastically reducing energy consumption, suitable for low-power embedded systems.

## Contribution

The paper presents the first integration of spiking neural networks with return-conditioned sequence modeling, combining biologically inspired mechanisms with decision transformers.

## Key findings

- Matches or exceeds standard Decision Transformer performance on control benchmarks
- Emits fewer than ten spikes per decision, indicating significant energy efficiency
- Potential for real-time, low-power control on embedded devices

## Abstract

Reinforcement learning agents based on Transformer architectures have achieved impressive performance on sequential decision-making tasks, but their reliance on dense matrix operations makes them ill-suited for energy-constrained, edge-oriented platforms. Spiking neural networks promise ultra-low-power, event-driven inference, yet no prior work has seamlessly merged spiking dynamics with return-conditioned sequence modeling. We present the Spiking Decision Transformer (SNN-DT), which embeds Leaky Integrate-and-Fire neurons into each self-attention block, trains end-to-end via surrogate gradients, and incorporates biologically inspired three-factor plasticity, phase-shifted spike-based positional encodings, and a lightweight dendritic routing module. Our implementation matches or exceeds standard Decision Transformer performance on classic control benchmarks (CartPole-v1, MountainCar-v0, Acrobot-v1, Pendulum-v1) while emitting fewer than ten spikes per decision, an energy proxy suggesting over four orders-of-magnitude reduction in per inference energy. By marrying sequence modeling with neuromorphic efficiency, SNN-DT opens a pathway toward real-time, low-power control on embedded and wearable devices.