相关论文: Constructing Multi-label Hierarchical Classificati…
We introduce the Cyber Defense Benchmark, a benchmark for measuring how well large language model (LLM) agents perform the core SOC analyst task of threat hunting: given a database of raw Windows event logs with no guided questions or…
Large-Scale Multi-Label Text Classification (LMTC) includes tasks with hierarchical label spaces, such as automatic assignment of ICD-9 codes to discharge summaries. Performance of models in prior art is evaluated with standard precision,…
Multi-agent systems extend large language models (LLMs) by decomposing tasks among specialized agents, but their distributed decision process creates new attack surfaces. We identify semantic hijacking, an attack in which harmful requests…
Air Traffic Control (ATC) is a safety-critical domain in which incorrect interpretation of instructions may lead to severe operational consequences. While large language models (LLMs) demonstrate strong general performance, their…
In this study, we implement a novel BERT architecture for multitask fine-tuning on three downstream tasks: sentiment classification, paraphrase detection, and semantic textual similarity prediction. Our model, Multitask BERT, incorporates…
Large language models (LLMs) have achieved impressive results across a range of natural language processing tasks, but their potential to generate harmful content has raised serious safety concerns. Current toxicity detectors primarily rely…
In this research, we use user defined labels from three internet text sources (Reddit, Stackexchange, Arxiv) to train 21 different machine learning models for the topic classification task of detecting cybersecurity discussions in natural…
Large language models (LLMs) remain vulnerable to sophisticated prompt engineering attacks that exploit contextual framing to bypass safety mechanisms, posing significant risks in cybersecurity applications. We introduce Jailbreak Mimicry,…
Semi-supervised text classification (SSTC) has gained increasing attention due to its ability to leverage unlabeled data. However, existing approaches based on pseudo-labeling suffer from the issues of pseudo-label bias and error…
Many domains now leverage the benefits of Machine Learning (ML), which promises solutions that can autonomously learn to solve complex tasks by training over some data. Unfortunately, in cyberthreat detection, high-quality data is hard to…
The increasingly sophisticated and growing number of threat actors along with the sheer speed at which cyber attacks unfold, make timely identification of attacks imperative to an organisations' security. Consequently, persons responsible…
Automatic topic classification has been studied extensively to assist managing and indexing scientific documents in a digital collection. With the large number of topics being available in recent years, it has become necessary to arrange…
The current state of Advanced Persistent Threats (APT) attribution primarily relies on time-consuming manual processes. These include mapping incident artifacts onto threat attribution frameworks and employing expert reasoning to uncover…
Public and commercial organizations extensively share cyberthreat intelligence (CTI) to prepare systems to defend against existing and emerging cyberattacks. However, traditional CTI has primarily focused on tracking known threat indicators…
Self-Admitted Technical Debt (SATD) refers to technical compromises explicitly admitted by developers in natural language artifacts such as code comments, commit messages, and issue trackers. Among its types, Architecture Technical Debt…
Assigning a set of labels to a given text is a classification problem with many real-world applications, such as recommender systems. Two separate research streams address this issue. Hierarchical Text Classification (HTC) focuses on…
Industrial multi-label document understanding pipelines score candidate labels and threshold or rank them to form a label set per document. This early selection step directly affects the accuracy of downstream information extraction from…
The Internet of Things (IoT) involves complex, interconnected systems and devices that depend on context-sharing platforms for interoperability and information exchange. These platforms are, therefore, critical components of real-world IoT…
Large language model-based web agents have shown strong potential in automating web interactions through advanced reasoning and instruction following. While retrieval-based memory derived from historical trajectories enables these agents to…
Intrusion Detection Systems (IDS) are critical to defending enterprise and industrial control environments, yet evaluating their effectiveness under realistic conditions remains an open challenge. Existing benchmarks rely on synthetic…