Optimal Attack

Optimal attack research focuses on designing and analyzing the most effective strategies for compromising machine learning systems and other complex systems, aiming to maximize the attacker's gain while minimizing detection. Current research employs diverse approaches, including reinforcement learning (e.g., Q-learning, multi-armed bandits), game theory (e.g., Stackelberg games), and neural network approximations of adversarial examples, to model and solve optimal attack problems across various contexts like federated learning and reinforcement learning agents. Understanding optimal attacks is crucial for developing robust defenses and ensuring the security and reliability of AI systems and other vulnerable technologies in diverse applications, from healthcare to finance.

Papers

November 1, 2024

Plentiful Jailbreaks with String Compositions
Brian R.Y. Huang
Jailbreak Attack Obfuscation Technique Word Level Attack Optimal Attack Invertible Linear Transformation World War II Era Cipher

October 22, 2024

Test-time Adversarial Defense with Opposite Adversarial Path and High Attack Time Cost
Cheng-Han Yeh, Kuanchun Yu, Chun-Shien Lu
Adversarial Attack Priority User Optimal Attack Adversarial Direction Test Time Adversarial

August 8, 2024

Constructing Adversarial Examples for Vertical Federated Learning: Optimal Client Corruption through Multi-Armed Bandit
Duanyi Yao, Songze Li, Ye Xue, Jin Liu
Adversarial Attack Adversarial Example Multi Armed Bandit Vertical Federated Learning Adversarial Corruption Optimal Attack

November 30, 2023

Optimal Attack and Defense for Reinforcement Learning
Jeremy McMahan, Young Wu, Xiaojin Zhu, Qiaomin Xie
Reinforcement Learning Stackelberg Game Adversarial Reinforcement Learning Defense Algorithm Observable Stochastic Game Optimal Attack

November 29, 2023

Q-learning Based Optimal False Data Injection Attack on Probabilistic Boolean Control Networks
Xianlun Peng, Yang Tang, Fangfei Li, Yang Liu
Reinforcement Learning Q Learning Optimal Attack Probabilistic Boolean Network

July 30, 2023

On Neural Network approximation of ideal adversarial attack and convergence of adversarial training
Rajdeep Haldar, Qifan Song
Adversarial Attack Adversarial Training Early Stage Convergence Adversarial Loss Learned Function Neural Network Approximation Optimal Attack

June 2, 2023

Hyperparameter Learning under Data Poisoning: Analysis of the Influence of Regularization via Multiobjective Bilevel Optimization
Javier Carnerero-Cano, Luis Muñoz-González, Phillippa Spencer, Emil C. Lupu
Native Robustness General Analysis Bilevel Optimization Data Poisoning Multi Step Attack Optimal Attack

May 8, 2023

Optimal Scheduling of Agents in ADTrees: Specialised Algorithm and Declarative Models
Jaime Arias, Carlos Olarte, Laure Petrucci, Łukasz Maśko, Wojciech Penczek, Teofil Sidoruk
Practical Algorithm Agent Smith Multi Agent Setting Optimal Scheduling Optimal Attack Declarative Model Attack Tree

February 27, 2023

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks
Mohammad Mohammadi, Jonathan Nöther, Debmalya Mandal, Adish Singla, Goran Radanovic
Poisoning Attack Training Time Attack Adversarial Policy Optimal Adversarial Optimal Attack

November 17, 2022

Towards Good Practices in Evaluating Transfer Adversarial Attacks
Zhengyu Zhao, Hanwei Zhang, Renjue Li, Ronan Sicre, Laurent Amsaleg, Michael Backes
Adversarial Attack Best Practice Transfer Attack Transfer Based Adversarial Attack Optimal Attack

October 15, 2022

A Scalable Reinforcement Learning Approach for Attack Allocation in Swarm to Swarm Engagement Problems
Umut Demir, Nazim Kemal Ure
Reinforcement Learning Swarm Control Optimal Attack Adversarial Swarm

April 25, 2022

Analysing the Influence of Attack Configurations on the Reconstruction of Medical Images in Federated Learning
Mads Emil Dahlgaard, Morten Wehlast Jørgensen, Niels Asp Fuglsang, Hiba Nassar
Medical Image Full State Reconstruction External Influence Target Distribution Private Training Optimal Attack Initialization Scheme

Optimal Attack

Papers

Plentiful Jailbreaks with String Compositions

Test-time Adversarial Defense with Opposite Adversarial Path and High Attack Time Cost

Constructing Adversarial Examples for Vertical Federated Learning: Optimal Client Corruption through Multi-Armed Bandit

Optimal Attack and Defense for Reinforcement Learning

Q-learning Based Optimal False Data Injection Attack on Probabilistic Boolean Control Networks

On Neural Network approximation of ideal adversarial attack and convergence of adversarial training

Hyperparameter Learning under Data Poisoning: Analysis of the Influence of Regularization via Multiobjective Bilevel Optimization

Optimal Scheduling of Agents in ADTrees: Specialised Algorithm and Declarative Models

Implicit Poisoning Attacks in Two-Agent Reinforcement Learning: Adversarial Policies for Training-Time Attacks

Towards Good Practices in Evaluating Transfer Adversarial Attacks

A Scalable Reinforcement Learning Approach for Attack Allocation in Swarm to Swarm Engagement Problems

Analysing the Influence of Attack Configurations on the Reconstruction of Medical Images in Federated Learning