-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy path_quarto.yml
More file actions
151 lines (142 loc) · 4.88 KB
/
Copy path_quarto.yml
File metadata and controls
151 lines (142 loc) · 4.88 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
project:
type: book
output-dir: docs
book:
output-file: thesis
title: "OverHAuL"
subtitle: "Harnessing Automation for C Libraries with Large Language Models"
date: 2025-07-27
author:
- name: "Konstantinos Chousos"
email: "kchousos@di.uoa.gr"
orcid: "0009-0008-6063-7915"
url: "https://kchousos.github.io/"
degrees:
- BSc
affiliations:
- name: "National and Kapodistrian University of Athens"
department: "Informatics and Telecommunications"
city: "Athens"
country: "Greece"
url: "https://en.uoa.gr/"
ror: "https://ror.org/04gnjpq42"
roles: writing
corresponding: true
license: CC BY
abstract: "Software vulnerabilities remain pervasive and challenging to detect, making robust testing approaches imperative. Fuzzing is an established software testing method for uncovering such vulnerabilities, through random input execution. Recent research has leveraged Large Language Models (LLMs) to enhance fuzz driver generation. However, most contemporary tools rely on additional resources beyond the target code, such as client programs or preexisting harnesses, limiting their scalability and applicability. In this thesis, we present OverHAuL, a neurosymbolic AI system that employs LLM agents to automatically generate fuzzing harnesses directly from library code, eliminating the need for auxiliary artifacts. To comprehensively evaluate OverHAuL, we construct a benchmark suite consisting of ten open-source C libraries. Our empirical analysis demonstrates that OverHAuL achieves an 81.25% success rate in harness generation across the evaluated projects, underscoring its effectiveness and potential to facilitate more efficient vulnerability discovery."
chapters:
- index.md
- chapters/intro.md
- chapters/background.md
- chapters/overhaul.md
- chapters/evaluation.md
- chapters/related.md
- chapters/future.md
- chapters/conclusion.md
- chapters/refs.md # Bibliography
appendices:
- chapters/abandoned.md
- chapters/sample-harnesses.md
- chapters/prompts.md
page-navigation: true
page-footer:
center:
- text: '© 2025 Konstantinos Chousos'
right:
- text: "{{< iconify fa6-brands github >}}"
url: "https://github.com/kchousos"
- text: "{{< iconify academicons orcid >}}"
url: "https://orcid.org/0009-0008-6063-7915"
- text: "{{< iconify fa6-brands linkedin >}}"
url: "https://www.linkedin.com/in/kchousos/"
- text: "{{< iconify material-symbols mail >}}"
url: "mailto:kchousos@di.uoa.gr"
downloads: [pdf]
repo-url: https://github.com/kchousos/BSc-Thesis/
repo-branch: master
# repo-actions: [edit, source]
bibliography: "./resources/bibliography.bib"
csl: "./resources/ieee-with-url.csl"
code-line-numbers: true
filters:
- pseudocode
format:
html:
default: true
html-math-method: katex
reference-location: block
citation: true
toc: true
toc-expand: true
number-sections: true
theme:
light:
- cosmo
- ./resources/styles.scss
dark:
- cosmo
- ./resources/styles.scss
- ./resources/styles-dark.scss
respect-user-color-scheme: true
highlight-style: arrow
date-format: long
lightbox: true
include-in-header:
text: |
<script src="https://tinylytics.app/embed/rSbhZ_sMtsVibUKrZtZ2/min.js" defer></script>
<script>
MathJax = {
loader: {
load: ['[tex]/boldsymbol']
},
tex: {
tags: "all",
inlineMath: [['$','$'], ['\\(','\\)']],
displayMath: [['$$','$$'], ['\\[','\\]']],
processEscapes: true,
processEnvironments: true,
packages: {
'[+]': ['boldsymbol']
}
}
};
</script>
<script src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-chtml-full.js" type="text/javascript"></script>
pdf:
default: true
date: 2025-07-27
date-format: MMMM, YYYY
documentclass: scrreprt
standalone: true
number-sections: true
keep-tex: true
toc: false
cite-method: biblatex
mainfont: "Libertinus Serif"
mainfontfallback:
- "Liberation Serif:"
- "NotoColorEmoji:mode=harf"
sansfont: "Libertinus Sans"
monofont: "Iosevka"
mathfont: "Libertinus Math"
numbersections: true
biblio-style: ieee
link-citations: true
papersize: a4
colorlinks: true
boxlinks: false
citecolor: red
filecolor: cyan
urlcolor: blue
links-as-notes: false
keywords: [LLMs, Fuzzing, Automation, Security, Neurosymbolic AI]
highlight-style: a11y
include-in-header: ./resources/preamble.tex
fig-scap: true
template-partials:
- ./resources/before-body.tex
pdf-engine: lualatex
pdf-engine-opts:
- '-shell-escape'
execute:
freeze: true