2026· Computers, Materials & Continua· pp. 1-10· 0 citations· 28 references
TL;DR
A seed optimization method called IRIS-Seed (Intelligent Reasoning and Integrated Scheduling for Path-Sensitive Seed Optimization) is presented, which consists of a chain-of-thought guided (CoT-guided) large language model (LLM) seed generation module, an execution-path collection module, and a path-sensitive seed selection and scheduling module.
Abstract
: Fuzzing has become one of the most widely used techniques in automated vulnerability discovery, with its core objective being to maximize code coverage by generating diverse test inputs. However, existing fuzzing techniques often generate low-quality seeds and overlook the diversity in seed execution paths, leading to suboptimal allocation of fuzzing resources. To address these issues, this paper presents a seed optimization method called IRIS-Seed (Intelligent Reasoning and Integrated Scheduling for Path-Sensitive Seed Optimization), which consists of three modules: a chain-of-thought guided (CoT-guided) large language model (LLM) seed generation module, an execution-path collection module, and a path-sensitive seed selection and scheduling module. The seed generation module employs an LLM as an engine for both input generation and expansion, using CoT prompting to produce diverse inputs tailored to the target programs. The seed selection module analyzes execution path coverage and constructs a filtering strategy based on path ordering relations, evaluating seed priority and assigning energy using metrics such as path similarity, path rarity, and edit distance, so that high-scoring seeds are scheduled preferentially. We integrated IRIS-Seed into three fuzzers and evaluated them on ten real-world programs. Overall, based on the aggregate results across the evaluated targets, IRIS-Seed discovered 1.18 times as many new edges and recorded 1.62 times as many saved crashes as the corresponding baseline fuzzers. Moreover, IRIS-Seed achieves better overall performance compared to other baseline methods.
This paper presents hybrid generator space fuzzing, a method that combines the strengths of hybrid fuzzing and generator-based fuzzing to improve the syntactic validity and semantic diversity of test inputs and implements GenEvo, an LLM-driven hybrid generator space fuzzer.
Si-Wei Wei, Shi-Hao Zhu, Rui-Jie Meng et al.· Proceedings of the 41st IEEE...· 1 citation
While coverage-guided kernel fuzzers have been proposed to uncover Linux kernel vulnerabilities, their code coverage and bug-finding capability are limited due to the lack of seed diversity, which is caused by the compounding effect of initial seed generation, seed scheduling, and seed mutation. To address this limitat...
Kun Hu, Jia-Ji Qin, Chao-Feng Sha et al.· Proceedings of the ACM on So...· 0 citations
Modern generator-based fuzzing techniques combine lightweight input generators with coverage-guided mutation as a method of exploring deep execution paths in a target program. A complementary approach in prior research focuses on creating highly customized, domain-specific generators that encode structural and semantic...
Vasudev Vikram, Rohan Padhye· Proceedings of the 41st IEEE...· 2 citations
Compiler fuzzing is critical for ensuring the reliability of modern optimizing compilers. However, traditional generators and raw Large Language Model (LLM) outputs often yield syntactically invalid or low-value programs, incurring significant computational overhead. In this paper, we present BanditFuzz, an end-to-end...
Dong-Hui Li, Ying-Ying Li, Jin-Long Xu et al.· Proceedings of the 41st IEEE...· 0 citations
Fuzzing finds bugs by testing software behavior with high volumes of input. To create semantically valid inputs for different targets, traditional fuzzers require significant engineering effort. LLM-based fuzzing approaches can be adapted to create inputs for a wide range of software but introduce new issues: every tes...
Lu Maltsis· Companion Proceedings of the...· 0 citations
Automated unit test generation promises to reduce the cost of software quality assurance, and hence, is attracting attention from both academia and industry. Yet, generating assertions that are executable, meaningful to developers, and able to catch faults remains an unsolved challenge. Existing approaches either rando...
Jia-Lun Cao, Hao-Yu Wang, Hao-Ran Yan et al.· Proceedings of the ACM on So...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.