Symmetry Defeats Auditing
密码学与安全
2026-05-28 v1 人工智能
摘要
We demonstrate an attack on Introspection Adapters (Shenoy et al., 2026).
引用
@article{arxiv.2605.27836,
title = {Symmetry Defeats Auditing},
author = {Nick Merrill and Zeke Medley},
journal= {arXiv preprint arXiv:2605.27836},
year = {2026}
}
相关论文
查看更多相关论文 →
密码学与安全 · 计算机科学
A Note on Implementation Errors in Recent Adaptive Attacks Against Multi-Resolution Self-Ensembles
Stanislav Fort
2025-01-27
密码学与安全 · 计算机科学
Revisiting the Robust Alignment of Circuit Breakers
Leo Schwinn, Simon Geisler
2024-08-05
机器学习 · 计算机科学
Symmetry Defense Against CNN Adversarial Perturbation Attacks
Blerta Lindqvist
2023-08-11
神经与进化计算 · 计算机科学
Adversarial Images for Variational Autoencoders
Pedro Tabacof, Julia Tavares, Eduardo Valle
2016-12-02
密码学与安全 · 计算机科学
One Step to the Side: Why Defenses Against Malicious Finetuning Fail Under Adaptive Adversaries
Itay Zloczower, Eyal Lenga, Gilad Gressel, Yisroel Mirsky
2026-05-26
机器学习 · 计算机科学
Encoding Involutory Invariances in Neural Networks
Anwesh Bhattacharya, Marios Mattheakis, Pavlos Protopapas
2022-04-28
chao-dyn · 物理学
Identification and Control of Symmetric Systems
R. O. Grigoriev
2007-05-23
计算机视觉与模式识别 · 计算机科学
Attacking Perceptual Similarity Metrics
Abhijay Ghildyal, Feng Liu
2023-05-16
宇宙学与河外天体物理 · 物理学
Lost in the Log-Polynomial Expansion: Comment on arXiv:2101.08278
Aritra Banerjee, Eoin Ó Colgáin, Misao Sasaki, M. M. Sheikh-Jabbari
2021-06-01
神经与进化计算 · 计算机科学
Symmetry Breaking in Neuroevolution: A Technical Report
Onay Urfalioglu, Orhan Arikan
2011-07-25
组合数学 · 数学
Projections, Entropy and Sumsets
Paul Balister, Béla Bollobás
2017-01-10
最优化与控制 · 数学
Symmetry reduction in AM/GM-based optimization
Philippe Moustrou, Helen Naumann, Cordian Riener, Thorsten Theobald +1
2021-12-09
机器学习 · 计算机科学
Hide & Seek: Transformer Symmetries Obscure Sharpness & Riemannian Geometry Finds It
Marvin F. da Silva, Felix Dangel, Sageev Oore
2025-05-09
数学物理 · 物理学
Symmetry problem
Alexander G. Ramm
2014-02-14
人工智能 · 计算机科学
Adaptive auditing of AI systems with anytime-valid guarantees
Siyu Zhou, Patrick Vossler, Venkatesh Sivaraman, Yifan Mai +1
2026-05-11
计算机视觉与模式识别 · 计算机科学
Cheating Stereo Matching in Full-scale: Physical Adversarial Attack against Binocular Depth Estimation in Autonomous Driving
Kangqiao Zhao, Shuo Huai, Xurui Song, Jun Luo
2026-03-17
经典分析与常微分方程 · 数学
On self-similar sets with overlaps and inverse theorems for entropy in $\mathbb{R}^d$
Michael Hochman
2017-06-07
密码学与安全 · 计算机科学
HBEE: Human Behavioral Entropy Engine -- Pre-Registered Multi-Agent LLM Simulation of Peer-Suspicion-Based Detection Inversion
Vickson Ferrel
2026-05-11
机器学习 · 计算机科学
Detecting Adversarial Examples Is (Nearly) As Hard As Classifying Them
Florian Tramèr
2022-06-17
人工智能 · 计算机科学
Beyond Mimicry: Toward Lifelong Adaptability in Imitation Learning
Nathan Gavenski, Felipe Meneguzzi, Odinaldo Rodrigues
2026-02-24
机器学习 · 计算机科学
On Adaptive Attacks to Adversarial Example Defenses
Florian Tramer, Nicholas Carlini, Wieland Brendel, Aleksander Madry
2020-10-26
计算复杂性 · 计算机科学
From expanders to hitting distributions and simulation theorems
Alexander Kozachinskiy
2018-07-09
信息论 · 计算机科学
Breaking the Bandwidth Barrier: Geometrical Adaptive Entropy Estimation
Weihao Gao, Sewoong Oh, Pramod Viswanath
2016-09-09
计算机科学与博弈论 · 计算机科学
Geometrical Regret Matching
Sizhong Lan
2020-01-24
计算机视觉与模式识别 · 计算机科学
See Eye to Eye: A Lidar-Agnostic 3D Detection Framework for Unsupervised Multi-Target Domain Adaptation
Darren Tsai, Julie Stephany Berrio, Mao Shan, Stewart Worrall +1
2023-04-12