Projects

A selection of my work I've done. My recent focus has been on spiking neural networks, biologically inspired learning, and reimplementing models from the literature to better understand their mechanics.

Projects

Spiking Neural Network Model of Hippocampal Reverse Replay

Dhruti Davey, Eleni Vasilaki

MSc Dissertation – University of Sheffield, 2024–2025

Reimplemented a reinforcement learning model of hippocampal reverse replay, originally rate based, and converted it to use spiking neurons. Place cells used Poisson encoding for spatial position, while action cells used leaky integrate-and-fire (LIF) neurons. The model simulates how an agent in a navigation task can replay trajectories in reverse order to accelerate learning, a mechanism observed biologically in the hippocampus. The spiking implementation preserved task performance while increasing biological plausibility.

[PDF] [Code]

computational neuroscience spiking neural networks hippocampus reinforcement learning

Publications

Large Language Model-Based Intrusion Detection System

Dhruti Davey, Kayvan Karim, Hani Ragab Hassen, Hadj Batatia

Springer Nature (published in a peer-reviewed volume), 2024

Developed an LLM-based intrusion detection system to classify network traffic as benign or malicious. Analysed NetFlow network data and logs, fine-tuned large language models for classification, and processed structured network traffic data to identify potential security threats.

[PDF] [Code]

cybersecurity large language models intrusion detection network security