Adversarially Pretrained Transformers may be Universally Robust In-Context Learners

20 May 2025

Soichiro Kumano

Hiroshi Kera

Toshihiko Yamasaki

Author Contacts:

kumano@cvm.t.u-tokyo.ac.jp kera@chiba-u.jp yamasaki@cvm.t.u-tokyo.ac.jp

AAML

ArXiv (abs)PDF HTML

Papers citing "Adversarially Pretrained Transformers may be Universally Robust In-Context Learners"

50 / 86 papers shown

Title
Evolution-based Region Adversarial Prompt Learning for Robustness Enhancement in Vision-Language Models Xiaojun Jia Sensen Gao Simeng Qin Ke Ma Xianrui Li Yihao Huang Wei Dong Yang Liu Xiaochun Cao AAML VLM 103 2 0 17 Mar 2025
On the Robustness of Transformers against Context Hijacking for Linear Classification Tianle Li Chenyang Zhang Xingwu Chen Yuan Cao Difan Zou 115 2 0 24 Feb 2025
Short-length Adversarial Training Helps LLMs Defend Long-length Jailbreak Attacks: Theoretical and Empirical Evidence Shaopeng Fu Liang Ding Di Wang Di Wang 102 4 0 06 Feb 2025
Adversarial Prompt Distillation for Vision-Language Models Lin Luo Xin Wang Bojia Zi Shihao Zhao Xingjun Ma Yu-Gang Jiang AAML VLM 135 4 0 22 Nov 2024
Adversarial Robustness of In-Context Learning in Transformers for Linear Regression Usman Anwar Johannes Von Oswald Louis Kirsch David M. Krueger Spencer Frei SILM AAML 44 9 0 07 Nov 2024
Can Looped Transformers Learn to Implement Multi-step Gradient Descent for In-context Learning? Khashayar Gatmiry Nikunj Saunshi Sashank J. Reddi Stefanie Jegelka Sanjiv Kumar 112 20 0 10 Oct 2024
Trained Transformer Classifiers Generalize and Exhibit Benign Overfitting In-Context Spencer Frei Gal Vardi MLT 63 6 0 02 Oct 2024
AdvQDet: Detecting Query-Based Adversarial Attacks with Adversarial Contrastive Prompt Tuning Xin Wang Kai-xiang Chen Xingjun Ma Zhineng Chen Jingjing Chen Yu-Gang Jiang AAML 95 4 0 04 Aug 2024
Why Larger Language Models Do In-context Learning Differently? Zhenmei Shi Junyi Wei Zhuoyan Xu Yingyu Liang 70 25 0 30 May 2024
Revisiting the Robust Generalization of Adversarial Prompt Tuning Fan Yang Mingxuan Xia Sangzhou Xia Chicheng Ma Hui Hui VPVLM VLM 42 3 0 18 May 2024
Few-Shot Adversarial Prompt Learning on Vision-Language Models Yiwei Zhou Xiaobo Xia Zhiwei Lin Bo Han Tongliang Liu VLM 85 16 0 21 Mar 2024
How Well Can Transformers Emulate In-context Newton's Method? Angeliki Giannou Liu Yang Tianhao Wang Dimitris Papailiopoulos Jason D. Lee 65 20 0 05 Mar 2024
One Prompt Word is Enough to Boost Adversarial Robustness for Pre-trained Vision-Language Models Lin Li Haoyan Guan Jianing Qiu Michael W. Spratling AAML VLM VPVLM 84 24 0 04 Mar 2024
Pre-trained Model Guided Fine-Tuning for Zero-Shot Adversarial Robustness Sibo Wang Jie Zhang Zheng Yuan Shiguang Shan VLM 60 24 0 09 Jan 2024
Transformers Implement Functional Gradient Descent to Learn Non-Linear Functions In Context Xiang Cheng Yuxin Chen S. Sra 49 41 0 11 Dec 2023
Scalable Extraction of Training Data from (Production) Language Models Milad Nasr Nicholas Carlini Jonathan Hayase Matthew Jagielski A. Feder Cooper Daphne Ippolito Christopher A. Choquette-Choo Eric Wallace Florian Tramèr Katherine Lee SILM 63 352 0 28 Nov 2023
Adversarial Prompt Tuning for Vision-Language Models Jiaming Zhang Xingjun Ma Xin Wang Lingyu Qiu Jiaqi Wang Yu-Gang Jiang Jitao Sang AAML VPVLM VLM 65 22 0 19 Nov 2023
Transformers as Decision Makers: Provable In-Context Reinforcement Learning via Supervised Pretraining Licong Lin Yu Bai Song Mei OffRL 72 48 0 12 Oct 2023
AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models Xiaogeng Liu Nan Xu Muhao Chen Chaowei Xiao SILM 77 324 0 03 Oct 2023
Uncovering mesa-optimization algorithms in Transformers J. Oswald Eyvind Niklasson Maximilian Schlegel Seijin Kobayashi Nicolas Zucchet ... Mark Sandler Blaise Agüera y Arcas Max Vladymyrov Razvan Pascanu João Sacramento 68 64 0 11 Sep 2023
Adversarial Finetuning with Latent Representation Constraint to Mitigate Accuracy-Robustness Tradeoff Satoshi Suzuki Shin'ya Yamaguchi Shoichiro Takeda Sekitoshi Kanai Naoki Makishima Atsushi Ando Ryo Masumura AAML 70 5 0 31 Aug 2023
Universal and Transferable Adversarial Attacks on Aligned Language Models Andy Zou Zifan Wang Nicholas Carlini Milad Nasr J. Zico Kolter Matt Fredrikson 291 1,455 0 27 Jul 2023
One Step of Gradient Descent is Provably the Optimal In-Context Learner with One Layer of Linear Self-Attention Arvind V. Mahankali Tatsunori B. Hashimoto Tengyu Ma MLT 63 101 0 07 Jul 2023
Jailbroken: How Does LLM Safety Training Fail? Alexander Wei Nika Haghtalab Jacob Steinhardt 200 970 0 05 Jul 2023
Supervised Pretraining Can Learn In-Context Reinforcement Learning Jonathan Lee Annie Xie Aldo Pacchiano Yash Chandak Chelsea Finn Ofir Nachum Emma Brunskill OffRL 96 82 0 26 Jun 2023
Trained Transformers Learn Linear Models In-Context Ruiqi Zhang Spencer Frei Peter L. Bartlett 79 201 0 16 Jun 2023
Transformers as Statisticians: Provable In-Context Learning with In-Context Algorithm Selection Yu Bai Fan Chen Haiquan Wang Caiming Xiong Song Mei 50 193 0 07 Jun 2023
Transformers learn to implement preconditioned gradient descent for in-context learning Kwangjun Ahn Xiang Cheng Hadi Daneshmand S. Sra ODL 86 170 0 01 Jun 2023
What and How does In-Context Learning Learn? Bayesian Model Averaging, Parameterization, and Generalization Yufeng Zhang Fengzhuo Zhang Zhuoran Yang Zhaoran Wang BDL 69 74 0 30 May 2023
Which Models have Perceptually-Aligned Gradients? An Explanation via Off-Manifold Robustness Suraj Srinivas Sebastian Bordt Hima Lakkaraju AAML 65 12 0 30 May 2023
The Learnability of In-Context Learning Noam Wies Yoav Levine Amnon Shashua 175 106 0 14 Mar 2023
Larger language models do in-context learning differently Jerry W. Wei Jason W. Wei Yi Tay Dustin Tran Albert Webson ... Xinyun Chen Hanxiao Liu Da Huang Denny Zhou Tengyu Ma ReLM LRM 97 372 0 07 Mar 2023
A Comprehensive Study on Robustness of Image Classification Models: Benchmarking and Rethinking Chang-Shu Liu Yinpeng Dong Wenzhao Xiang Xiaohu Yang Hang Su Junyi Zhu YueFeng Chen Yuan He H. Xue Shibao Zheng OOD VLM AAML 83 79 0 28 Feb 2023
Large Language Models Can Be Easily Distracted by Irrelevant Context Freda Shi Xinyun Chen Kanishka Misra Nathan Scales David Dohan Ed H. Chi Nathanael Scharli Denny Zhou ReLM RALM LRM 101 587 0 31 Jan 2023
Why Can GPT Learn In-Context? Language Models Implicitly Perform Gradient Descent as Meta-Optimizers Damai Dai Yutao Sun Li Dong Y. Hao Shuming Ma Zhifang Sui Furu Wei LRM 76 168 0 20 Dec 2022
Transformers learn in-context by gradient descent J. Oswald Eyvind Niklasson E. Randazzo João Sacramento A. Mordvintsev A. Zhmoginov Max Vladymyrov MLT 110 489 0 15 Dec 2022
Understanding Zero-Shot Adversarial Robustness for Large-Scale Models Chengzhi Mao Scott Geng Junfeng Yang Xin Eric Wang Carl Vondrick VLM 84 69 0 14 Dec 2022
What learning algorithm is in-context learning? Investigations with linear models Ekin Akyürek Dale Schuurmans Jacob Andreas Tengyu Ma Denny Zhou 100 487 0 28 Nov 2022
Ignore Previous Prompt: Attack Techniques For Language Models Fábio Perez Ian Ribeiro SILM 99 438 0 17 Nov 2022
A Light Recipe to Train Robust Vision Transformers Edoardo Debenedetti Vikash Sehwag Prateek Mittal ViT 76 71 0 15 Sep 2022
What Can Transformers Learn In-Context? A Case Study of Simple Function Classes Shivam Garg Dimitris Tsipras Percy Liang Gregory Valiant 139 506 0 01 Aug 2022
Towards Efficient Adversarial Training on Vision Transformers Boxi Wu Jindong Gu Zhifeng Li Deng Cai Xiaofei He Wei Liu ViT AAML 72 39 0 21 Jul 2022
Quantifying Memorization Across Neural Language Models Nicholas Carlini Daphne Ippolito Matthew Jagielski Katherine Lee Florian Tramèr Chiyuan Zhang PILM 121 623 0 15 Feb 2022
Are Transformers More Robust Than CNNs? Yutong Bai Jieru Mei Alan Yuille Cihang Xie ViT AAML 244 263 0 10 Nov 2021
Adversarial Robustness Comparison of Vision Transformer and MLP-Mixer to CNNs Philipp Benz Soomin Ham Chaoning Zhang Adil Karjauv In So Kweon AAML ViT 85 79 0 06 Oct 2021
RobustART: Benchmarking Robustness on Architecture Design and Training Techniques Shiyu Tang Ruihao Gong Yan Wang Aishan Liu Jiakai Wang ... Xianglong Liu Basel Alomair Alan Yuille Philip Torr Dacheng Tao VLM AAML 41 108 0 11 Sep 2021
Long-term Cross Adversarial Training: A Robust Meta-learning Method for Few-shot Classification Tasks Fan Liu Shuyu Zhao Xuelong Dai Bin Xiao VLM 63 8 0 22 Jun 2021
Reveal of Vision Transformers Robustness against Adversarial Attacks Ahmed Aldahdooh W. Hamidouche Olivier Déforges ViT 43 60 0 07 Jun 2021
Intriguing Properties of Vision Transformers Muzammal Naseer Kanchana Ranasinghe Salman Khan Munawar Hayat Fahad Shahbaz Khan Ming-Hsuan Yang ViT 321 649 0 21 May 2021
Vision Transformers are Robust Learners Sayak Paul Pin-Yu Chen ViT 61 311 0 17 May 2021