CASCADE: LLM-Powered JavaScript Deobfuscator at Google
Not on the map. Why?
Abstract (arXiv)
Software obfuscation, particularly prevalent in JavaScript, hinders code comprehension and analysis, posing significant challenges to software testing, static analysis, and malware detection. This paper introduces CASCADE, a novel hybrid approach that integrates the advanced coding capabilities of Gemini with the deterministic transformation capabilities of a compiler Intermediate Representation (IR), specifically JavaScript IR (JSIR). By employing Gemini to identify critical prelude functions, the foundational components underlying the most prevalent obfuscation techniques, and leveraging JSIR for subsequent code transformations, CASCADE effectively recovers semantic elements like original strings and API names, and reveals original program behaviors. This method overcomes limitations of existing static and dynamic deobfuscation techniques, eliminating hundreds to thousands of hardcoded rules while achieving reliability and flexibility. CASCADE is already deployed in Google's production environment, demonstrating substantial improvements in JavaScript deobfuscation efficiency and reducing reverse engineering efforts.
Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 13 of its 55 references that are in the map. First posted on arXiv 23 Jul 2025.
Builds on
13 of 55 references.
References · 13
- 2023Are Emergent Abilities of Large Language Models a Mirage?cited by 523
- 2024StarCoder 2 and The Stack v2: The Next Generationcited by 502
- 2023Using an LLM to Help With Code Understandingcited by 228
- 2023How well do Large Language Models perform in Arithmetic tasks?cited by 130
- 2022Code Translation with Compiler Representationscited by 42
- 2022CCTEST: Testing and Repairing Code Completion Systemscited by 35
- 2024Language Models Do Hard Arithmetic Tasks Easily and Hardly Do Easy Arithmetic Taskscited by 24
- 2024LLM-Powered Test Case Generation for Detecting Bugs in Plausible Programscited by 14
- 2025On the Effectiveness of Large Language Models in Writing Alloy Formulascited by 7
- 2025JsDeObsBench: Measuring and Benchmarking LLMs for JavaScript Deobfuscationcited by 7
- 2024Generating executable oracles to check conformance of client code to requirements of JDK Javadocs using LLMscited by 6
- 2024FV8: A Forced Execution JavaScript Engine for Detecting Evasive Techniquescited by 3
1 more
- 2020Optimizing Away JavaScript Obfuscationcited by 2
Built on by
Citing papers per year: 2025: 2; 2026: 3.
5 papers.
- 2025OBsmith: LLM-Powered JavaScript Obfuscator Testingcited by 2
- 2026Semantic Voting: Execution-Grounded Consensus for LLM Code Generationcited by 1
- 2025APRIL: API Synthesis with Automatic Prompt Optimization and Reinforcement Learningcited by 0
- 2026Poisoned Identifiers Survive LLM Deobfuscation: A Case Study on Claude Opus 4.6cited by 0
- 2026Sketch-and-Verify: Structured Inference-Time Scaling via Program Sketchingcited by 0
- 2026Semantic Voting: Execution-Grounded Consensus for LLM Code Generation9 May 2026
- 2026Sketch-and-Verify: Structured Inference-Time Scaling via Program Sketching9 May 2026
- 2026Poisoned Identifiers Survive LLM Deobfuscation: A Case Study on Claude Opus 4.65 Apr 2026
- 2025OBsmith: LLM-Powered JavaScript Obfuscator Testing11 Oct 2025
- 2025APRIL: API Synthesis with Automatic Prompt Optimization and Reinforcement Learning29 Aug 2025