Files and links (1)
Published, Version of Record (VoR)Open Access via Drexel Libraries Read and Publish Program 2025CC BY V4.0, Open
Conference proceeding
LIFT: Automating Symbolic Execution Optimization with Large Language Models for AI Networks
NAIC '25: Proceedings of the 2nd Workshop on Networks for AI Computing, pp 37-42
08 Sep 2025
Featured in Collection : Research Supported by Drexel Libraries' OA Programs
Abstract
Dynamic Symbolic Execution (DSE) is a key technique in program analysis, widely used in software testing, vulnerability discovery, and formal verification. In distributed AI systems, DSE plays a crucial role in identifying hard-to-detect bugs, especially those arising from complex network communication patterns. However, traditional approaches to symbolic execution are often hindered by scalability issues and inefficiencies, particularly in large-scale systems. This paper introduces LIFT (Large-language-model Integrated Functional-equivalent-IR Transformation), a novel framework that leverages Large Language Models (LLMs) to automate the optimization of Intermediate Representations (IRs) in symbolic execution. LIFT addresses the challenges of symbolic execution by providing a scalable, context-sensitive solution for IR transformation. The framework consists of two phases: IR Analysis and Optimization, where LLMs optimize time-intensive IR blocks, and Symbolic Execution and Validation, which includes benchmarking and semantic verification to ensure correctness and generalizability. Experiments on real-world binaries demonstrated significant performance improvements, including a 53.5% reduction in execution time for bigtest and a 10.24% reduction for random, along with reductions in IR statements, PUT instructions, and temporary variables. These results demonstrate that LLMs simplify IRs while maintaining functional correctness, enhancing symbolic execution in distributed AI systems.
Metrics
12 Record Views
Details
- Title
- LIFT: Automating Symbolic Execution Optimization with Large Language Models for AI Networks
- Creators
- Ruoxi Wang - Northeastern UniversityKun Li - Shandong University of Science and TechnologyMinghui Xu - Shandong University of Science and TechnologyYue Zhang - Drexel University, Computer ScienceKaidi Xu (Corresponding Author) - Drexel University, Computer ScienceChunchi LiuYinhao Xiao - Guangdong University Of Finances and EconomicsXiuzhen Cheng - Shandong University
- Publication Details
- NAIC '25: Proceedings of the 2nd Workshop on Networks for AI Computing, pp 37-42
- Publisher
- Association for Computing Machinery
- Number of pages
- 6
- Grant note
- Key R&D Program (Soft Science Project) of Shandong Province, China: 2024RZB0406 National Natural Science Foundation of China: 62302266, 62232010, U23A20302, U24A20244 Shandong Science Fund for Excellent Young Scholars: ZR2022ZD02 Shandong Provincial Natural Science FoundationGuangdong Basic and Applied Basic Research Foundation: 2025A1515010111
This study was supported by Key R&D Program (Soft Science Project) of Shandong Province, China (2024RZB0406), the National Natural Science Foundation of China (No. 62302266, 62232010, U23A20302, U24A20244), the Shandong Science Fund for Excellent Young Scholars (No.2023HWYQ-008), the project ZR2022ZD02 supported by Shandong Provincial Natural Science Foundation and the Guangdong Basic and Applied Basic Research Foundation (Grant No. 2025A1515010111).
- Resource Type
- Conference proceeding
- Language
- English
- Academic Unit
- Computer Science
- Web of Science ID
- WOS:001592390500007
- Scopus ID
- 2-s2.0-105018059012
- Other Identifier
- 991022085455804721
InCites Highlights
Data related to this publication, from InCites Benchmarking & Analytics tool:
- Collaboration types
- Industry collaboration
- Domestic collaboration
- International collaboration
- Web of Science research areas
- Computer Science, Artificial Intelligence
- Telecommunications