Large Language Model-Based Reward Design for Deep Reinforcement Learning-Driven Autonomous Cyber Defense

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mukherjee, Sayak, Chatterjee, Samrat, Purvine, Emilie, Fujimoto, Ted, Emerson, Tegan
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914165708292096
author Mukherjee, Sayak
Chatterjee, Samrat
Purvine, Emilie
Fujimoto, Ted
Emerson, Tegan
author_facet Mukherjee, Sayak
Chatterjee, Samrat
Purvine, Emilie
Fujimoto, Ted
Emerson, Tegan
contents Designing rewards for autonomous cyber attack and defense learning agents in a complex, dynamic environment is a challenging task for subject matter experts. We propose a large language model (LLM)-based reward design approach to generate autonomous cyber defense policies in a deep reinforcement learning (DRL)-driven experimental simulation environment. Multiple attack and defense agent personas were crafted, reflecting heterogeneity in agent actions, to generate LLM-guided reward designs where the LLM was first provided with contextual cyber simulation environment information. These reward structures were then utilized within a DRL-driven attack-defense simulation environment to learn an ensemble of cyber defense policies. Our results suggest that LLM-guided reward designs can lead to effective defense strategies against diverse adversarial behaviors.
format Preprint
id arxiv_https___arxiv_org_abs_2511_16483
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Large Language Model-Based Reward Design for Deep Reinforcement Learning-Driven Autonomous Cyber Defense
Mukherjee, Sayak
Chatterjee, Samrat
Purvine, Emilie
Fujimoto, Ted
Emerson, Tegan
Machine Learning
Artificial Intelligence
Multiagent Systems
Designing rewards for autonomous cyber attack and defense learning agents in a complex, dynamic environment is a challenging task for subject matter experts. We propose a large language model (LLM)-based reward design approach to generate autonomous cyber defense policies in a deep reinforcement learning (DRL)-driven experimental simulation environment. Multiple attack and defense agent personas were crafted, reflecting heterogeneity in agent actions, to generate LLM-guided reward designs where the LLM was first provided with contextual cyber simulation environment information. These reward structures were then utilized within a DRL-driven attack-defense simulation environment to learn an ensemble of cyber defense policies. Our results suggest that LLM-guided reward designs can lead to effective defense strategies against diverse adversarial behaviors.
title Large Language Model-Based Reward Design for Deep Reinforcement Learning-Driven Autonomous Cyber Defense
topic Machine Learning
Artificial Intelligence
Multiagent Systems
url https://arxiv.org/abs/2511.16483