Adversarial Scenario

Adversarial scenarios in machine learning explore how models behave under malicious attacks, aiming to understand and mitigate vulnerabilities. Current research focuses on developing robust models and algorithms, including adversarial training techniques, uncertainty quantification methods, and the use of ensemble methods, often applied to neural networks, transformers, and reinforcement learning agents. This field is crucial for ensuring the reliability and safety of AI systems across various applications, from autonomous vehicles and network security to healthcare and language models, where the consequences of model failure can be severe. The ultimate goal is to build more trustworthy and resilient AI systems capable of operating effectively in unpredictable and potentially hostile environments.

Papers

October 21, 2024

October 15, 2024

Taking off the Rose-Tinted Glasses: A Critical Look at Adversarial ML Through the Lens of Evasion Attacks
Kevin Eykholt, Farhan Ahmed, Pratik Vaishnavi, Amir Rahmati
Adversarial Attack Adversarial Learning Camera Lens Critical Review Adversarial Scenario Evasion Attack Adversarial Threat Adversarial Evasion Attack

September 12, 2024

Alignment with Preference Optimization Is All You Need for LLM Safety
Reda Alami, Ali Khalifa Almansoori, Ahmed Alzubaidi, Mohamed El Amine Seddik, Mugariya Farooq, Hakim Hacid
Alignment Problem Preference Optimization Robust Model LLM Safety Adversarial Scenario Accident Datasets

September 8, 2024

September 3, 2024

Dynamic Guidance Adversarial Distillation with Enhanced Teacher Knowledge
Hyejin Park, Dongbo Min
Adversarial Input Adversarial Scenario Adversarial Threat Adversarial Distillation

August 27, 2024

Adversarial Manhole: Challenging Monocular Depth Estimation and Semantic Segmentation Models with Patch Attack
Naufal Suryanto, Andro Aprila Adiputra, Ahmada Yusril Kadiptya, Yongsu Kim, Howon Kim
Semantic Segmentation Monocular Depth Estimation Depth Map Semantic Segmentation Model Adversarial Patch Adversarial Scenario Patch Attack Adversarial Backdoor

August 25, 2024

On the Robustness of Kolmogorov-Arnold Networks: An Adversarial Perspective
Tal Alter, Raz Lapid, Moshe Sipper
Native Robustness Adversarial Example Adversarial Robustness Kolmogorov Arnold Network Adversarial Scenario Adversarial Perspective White Box Adversarial Attack

August 21, 2024

CSPI-MT: Calibrated Safe Policy Improvement with Multiple Testing for Threshold Policies
Brian M Cho, Ana-Roxana Pop, Kyra Gan, Sam Corbett-Davies, Israel Nir, Ariel Evnine, Nathan Kallus
Adversarial Scenario Policy Improvement Multiple Hypothesis Testing Optimal Safe Policy

August 20, 2024

Learning Randomized Algorithms with Transformers
Johannes von Oswald, Seijin Kobayashi, Yassir Akram, Angelika Steger
Transformer Megatron Decepticons Adversarial Scenario Adversarial Objective Randomization Program Randomized Algorithm Oblivious Adversary

August 6, 2024

Adversarial Safety-Critical Scenario Generation using Naturalistic Human Driving Priors
Kunkun Hao, Yonggang Luo, Wen Cui, Yuqiao Bai, Jucheng Yang, Songyang Yan, Yuxi Pan, Zijiang Yang
Autonomous Vehicle Adversarial Scenario Naturalistic Driving Intelligent Driver Model Generative Adversarial Imitation

July 22, 2024

On Feasibility of Intent Obfuscating Attacks
Zhaobin Li, Patrick Shafto
Adversarial Attack Adversarial Example Task Feasibility Adversarial Scenario

June 24, 2024

UNICAD: A Unified Approach for Attack Detection, Noise Reduction and Novel Class Identification
Alvaro Lopez Pellicer, Kittipos Giatgong, Yi Li, Neeraj Suri, Plamen Angelov
Adversarial Attack Unified Framework Adversarial DEfense Unseen Class Novel Class Discovery Adversarial Scenario Noise Reduction Attack Detection

June 5, 2024

FREA: Feasibility-Guided Generation of Safety-Critical Scenarios with Reasonable Adversariality
Keyu Chen, Yuheng Lei, Hao Cheng, Haoran Wu, Wenchao Sun, Sifa Zheng
Autonomous Vehicle Adversarial Scenario Safety Critical Scenario Adversarial Policy Critical Scenario Generation

June 1, 2024

LIDAO: Towards Limited Interventions for Debiasing (Large) Language Models
Tianci Liu, Haoyu Wang, Shiyang Wang, Yu Cheng, Jing Gao
Large Language Model Language Model Procedural Fairness Language Generation Self Debiasing Adversarial Scenario Fairness Aware

May 2, 2024

Error-Driven Uncertainty Aware Training
Pedro Mendes, Paolo Romano, David Garlan
Uncertainty Estimation Model Uncertainty Adversarial Scenario Uncertainty Aware Training

April 21, 2024

Reliable Model Watermarking: Defending Against Theft without Compromising on Evasion
Hongyu Zhu, Sichu Liang, Wentao Hu, Fangqi Li, Ju Jia, Shilin Wang
Watermarking Method Adversarial Scenario Model Watermarking Unrestricted Adversarial

March 3, 2024

GuardT2I: Defending Text-to-Image Models from Adversarial Prompts
Yijun Yang, Ruiyuan Gao, Xiao Yang, Jianyuan Zhong, Qiang Xu
Text to Image Model Text to Image Adversarial Prompt Adversarial Scenario Textual Guidance

February 20, 2024

Bounding Reconstruction Attack Success of Adversaries Without Data Priors
Alexander Ziller, Anneliese Riess, Kristian Schwethelm, Tamara T. Mueller, Daniel Rueckert, Georgios Kaissis
Differential Privacy Sensitive Data Adversarial Scenario Upper Bound Adversary Agent Reconstruction Attack Data Driven Prior