Image de couverture de livre

PISmith: Reinforcement Learning-based Red Teaming for Prompt Injection Defenses

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Yin, Chenlong, Geng, Runpeng, Wang, Yanting, Jia, Jinyuan
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning Cryptography and Security
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

PIArena: A Platform for Prompt Injection Evaluation
par: Geng, Runpeng, et autres
Publié: (2026)

PISanitizer: Preventing Prompt Injection to Long-Context LLMs via Prompt Sanitization
par: Geng, Runpeng, et autres
Publié: (2025)

FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption
par: Wang, Yanting, et autres
Publié: (2026)

Formalizing and Benchmarking Prompt Injection Attacks and Defenses
par: Liu, Yupei, et autres
Publié: (2023)

AgentWatcher: A Rule-based Prompt Injection Monitor
par: Wang, Yanting, et autres
Publié: (2026)

AttnTrace: Contextual Attribution of Prompt Injection and Knowledge Corruption
par: Wang, Yanting, et autres
Publié: (2025)

TracLLM: A Generic Framework for Attributing Long Context LLMs
par: Wang, Yanting, et autres
Publié: (2025)

PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models
par: Zou, Wei, et autres
Publié: (2024)

PIShield: Detecting Prompt Injection Attacks via Intrinsic LLM Features
par: Zou, Wei, et autres
Publié: (2025)

TASO: Jailbreak LLMs via Alternative Template and Suffix Optimization
par: Wang, Yanting, et autres
Publié: (2025)

UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented Generation
par: Geng, Runpeng, et autres
Publié: (2025)

A Multi-Agent LLM Defense Pipeline Against Prompt Injection Attacks
par: Hossain, S M Asif, et autres
Publié: (2025)

AttackGNN: Red-Teaming GNNs in Hardware Security Using Reinforcement Learning
par: Gohil, Vasudev, et autres
Publié: (2024)

RedVisor: Reasoning-Aware Prompt Injection Defense via Zero-Copy KV Cache Reuse
par: Liu, Mingrui, et autres
Publié: (2026)

Adaptive Attacks Break Defenses Against Indirect Prompt Injection Attacks on LLM Agents
par: Zhan, Qiusi, et autres
Publié: (2025)

AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
par: Debenedetti, Edoardo, et autres
Publié: (2024)

Send a SCOUT First: Pre-hoc Reasoning for Adaptive Detector Allocation in Prompt-Injection Defense
par: Zhang, Shuhao, et autres
Publié: (2026)

The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections
par: Nasr, Milad, et autres
Publié: (2025)

SecInfer: Preventing Prompt Injection via Inference-time Scaling
par: Liu, Yupei, et autres
Publié: (2025)

Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation
par: Morasso, Cristian, et autres
Publié: (2026)

Leveraging Reinforcement Learning in Red Teaming for Advanced Ransomware Attack Simulations
par: Wang, Cheng, et autres
Publié: (2024)

RL Is a Hammer and LLMs Are Nails: A Simple Reinforcement Learning Recipe for Strong Prompt Injection
par: Wen, Yuxin, et autres
Publié: (2025)

Automated Cyber Defense with Generalizable Graph-based Reinforcement Learning Agents
par: King, Isaiah J., et autres
Publié: (2025)

Early Approaches to Adversarial Fine-Tuning for Prompt Injection Defense: A 2022 Study of GPT-3 and Contemporary Models
par: Sandoval, Gustavo, et autres
Publié: (2025)

A Critical Evaluation of Defenses against Prompt Injection Attacks
par: Jia, Yuqi, et autres
Publié: (2025)

EnsembleSHAP: Faithful and Certifiably Robust Attribution for Random Subspace Method
par: Wang, Yanting, et autres
Publié: (2026)

LeakSealer: A Semisupervised Defense for LLMs Against Prompt Injection and Leakage Attacks
par: Panebianco, Francesco, et autres
Publié: (2025)

Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks
par: Pasquini, Dario, et autres
Publié: (2024)

Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs
par: Maiorano, Alexandre Cristovão
Publié: (2026)

Preventing Prompt Injection with Type-Directed Privilege Separation
par: Jacob, Dennis, et autres
Publié: (2025)

Co-RedTeam: Orchestrated Security Discovery and Exploitation with LLM Agents
par: He, Pengfei, et autres
Publié: (2026)

Red Teaming with Artificial Intelligence-Driven Cyberattacks: A Scoping Review
par: Al-Azzawi, Mays, et autres
Publié: (2025)

SecAlign: Defending Against Prompt Injection with Preference Optimization
par: Chen, Sizhe, et autres
Publié: (2024)

Design Patterns for Securing LLM Agents against Prompt Injections
par: Beurer-Kellner, Luca, et autres
Publié: (2025)

Lessons from Defending Gemini Against Indirect Prompt Injections
par: Shi, Chongyang, et autres
Publié: (2025)

Checkpoint-GCG: Auditing and Attacking Fine-Tuning-Based Prompt Injection Defenses
par: Yang, Xiaoxue, et autres
Publié: (2025)

Mitigating Indirect Prompt Injection via Instruction-Following Intent Analysis
par: Kang, Mintong, et autres
Publié: (2025)

From Firewalls to Frontiers: AI Red-Teaming is a Domain-Specific Evolution of Cyber Red-Teaming
par: Sinha, Anusha, et autres
Publié: (2025)

Evaluating Differential Privacy Against Membership Inference in Federated Learning: Insights from the NIST Genomics Red Team Challenge
par: Bertoli, Gustavo de Carvalho
Publié: (2026)

Noise as a Double-Edged Sword: Reinforcement Learning Exploits Randomized Defenses in Neural Networks
par: Bakos, Steve, et autres
Publié: (2024)