MétaCan
Menu
Back to cohort

A Generic Blue Agent Training Framework for Autonomous Cyber Operations

2024· article· en· W4401597644 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

Venuenot available
Typearticle
Languageen
FieldComputer Science
TopicMobile Agent-Based Network Management
Canadian institutionsCarleton University
Fundersnot available
KeywordsComputer scienceTraining (meteorology)Computer security

Abstract

fetched live from OpenAlex

Sophisticated mechanisms for attacking a computer network are emerging, therefore it is of great importance that equally sophisticated mechanisms should be in place to defend against malicious attacks on the network. Autonomous cyber operations (ACO) is considered to be a potential option to provide timely defense against malicious attacks. In ACO, an agent that tries to attack a network is referred to as red agent, and an agent that defends against the red agent is called blue agent. In real scenarios, different kinds of red agents can attack a network, hence a blue agent needs to defend against a variety of red agents, each with their own attack strategy and specific goal. However, it is a challenging task to train a blue agent that is agnostic of the red agent. Hence, we present here a framework for generic blue agent training, i.e., training a blue agent that can defend against different kinds of red agents. The framework is a combination of reinforcement learning and supervised learning. Our results demonstrate that the presented framework for generic blue agent training does exhibit generic characteristics, and the framework does demonstrate better performance compared to an alternate approach.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: Methods
Teacher disagreement score0.737
Threshold uncertainty score0.670

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0010.000
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.048
GPT teacher head0.286
Teacher spread0.238 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Quick stats

Citations5
Published2024
Admission routes1
Has abstractyes

Explore more

Same topicMobile Agent-Based Network ManagementFrench-language works237,207