LocalAlign: Enabling Generalizable Prompt Injection Defense via Generation of Near-Target Adversarial Examples for Alignment Training
Open Access
- 14 November 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 5388-5402
Abstract
No abstract availableThis publication has 15 references indexed in Scilit:
- Jailbreak and Guard Aligned Language Models With Only Few In-Context DemonstrationsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2026
- FlippedRAG: Black-Box Opinion Manipulation Adversarial Attacks to Retrieval-Augmented Generation Models2025
- SecAlign: Defending Against Prompt Injection with Preference Optimization2025
- A Survey on Trustworthy LLM Agents: Threats and Countermeasures2025
- DataSentinel: A Game-Theoretic Detection of Prompt Injection Attacks2025
- "Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models2024
- Large Language Models Are Involuntary Truth-Tellers: Exploiting Fallacy Failure for Jailbreak Attacks2024
- Not What You've Signed Up For: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection2023
- Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them2023
- A Dataset of Information-Seeking Questions and Answers Anchored in Research Papers2021