English
Related papers

Related papers: Winter Soldier: Backdooring Language Models at Pre…

200 papers

Poisoning of data sets is a potential security threat to large language models that can lead to backdoored models. A description of the internal mechanisms of backdoored language models and how they process trigger inputs, e.g., when…

Machine Learning · Computer Science 2024-05-07 Max Lamparth , Anka Reuel

We introduce camouflaged data poisoning attacks, a new attack vector that arises in the context of machine unlearning and other settings when model retraining may be induced. An adversary first adds a few carefully crafted points to the…

Machine Learning · Computer Science 2024-08-02 Jimmy Z. Di , Jack Douglas , Jayadev Acharya , Gautam Kamath , Ayush Sekhari

Detecting semantic backdoors in classification models--where some classes can be activated by certain natural, but out-of-distribution inputs--is an important problem that has received relatively little attention. Semantic backdoors are…

Machine Learning · Computer Science 2026-01-08 Arpad Berta , Gabor Danner , Istvan Hegedus , Mark Jelasity

Large language models (LLMs) are excellent in-context learners. However, the sensitivity of data contained in prompts raises privacy concerns. Our work first shows that these concerns are valid: we instantiate a simple but highly effective…

Machine Learning · Computer Science 2023-05-26 Haonan Duan , Adam Dziedzic , Nicolas Papernot , Franziska Boenisch

Large pre-trained language models (PLMs) have proven to be a crucial component of modern natural language processing systems. PLMs typically need to be fine-tuned on task-specific downstream datasets, which makes it hard to claim the…

Computation and Language · Computer Science 2023-02-13 Chenxi Gu , Chengsong Huang , Xiaoqing Zheng , Kai-Wei Chang , Cho-Jui Hsieh

Large language models (LLMs) are increasingly deployed in settings where inducing a bias toward a certain topic can have significant consequences, and backdoor attacks can be used to produce such models. Prior work on backdoor attacks has…

Cryptography and Security · Computer Science 2026-02-17 Anudeep Das , Prach Chantasantitam , Gurjot Singh , Lipeng He , Mariia Ponomarenko , Florian Kerschbaum

Substantial research works have shown that deep models, e.g., pre-trained models, on the large corpus can learn universal language representations, which are beneficial for downstream NLP tasks. However, these powerful models are also…

Cryptography and Security · Computer Science 2024-07-16 Yixin Liu , Hongsheng Hu , Xun Chen , Xuyun Zhang , Lichao Sun

System Instructions in Large Language Models (LLMs) are commonly used to enforce safety policies, define agent behavior, and protect sensitive operational context in agentic AI applications. These instructions may contain sensitive…

Cryptography and Security · Computer Science 2026-04-02 Anubhab Sahu , Diptisha Samanta , Reza Soosahabi

Pretraining language models directly on web-scale corpora is the de facto paradigm. We study an alternative where the model is initially exposed to abstract structured data to ease the subsequent acquisition of rich semantic knowledge, much…

Computation and Language · Computer Science 2026-05-29 Liangze Jiang , Zachary Shinnick , Anton van den Hengel , Hemanth Saratchandran , Damien Teney

Current techniques for privacy auditing of large language models (LLMs) have limited efficacy -- they rely on basic approaches to generate canaries which leads to weak membership inference attacks that in turn give loose lower bounds on the…

Cryptography and Security · Computer Science 2025-03-11 Ashwinee Panda , Xinyu Tang , Milad Nasr , Christopher A. Choquette-Choo , Prateek Mittal

Large Language Models (LLMs), while powerful, are built and trained to process a single text input. In common applications, multiple inputs can be processed by concatenating them together into a single stream of text. However, the LLM is…

Cryptography and Security · Computer Science 2024-03-25 Keegan Hines , Gary Lopez , Matthew Hall , Federico Zarfati , Yonatan Zunger , Emre Kiciman

Large language models (LLMs) are vulnerable when trained on datasets containing harmful content, which leads to potential jailbreaking attacks in two scenarios: the integration of harmful texts within crowdsourced data used for pre-training…

Cryptography and Security · Computer Science 2024-06-03 Xiaoqun Liu , Jiacheng Liang , Muchao Ye , Zhaohan Xi

Deep neural networks are vulnerable to adversarial attacks, such as backdoor attacks in which a malicious adversary compromises a model during training such that specific behaviour can be triggered at test time by attaching a specific word…

Cryptography and Security · Computer Science 2022-10-21 You Guo , Jun Wang , Trevor Cohn

Natural language processing models have experienced a significant upsurge in recent years, with numerous applications being built upon them. Many of these applications require fine-tuning generic base models on customized, proprietary…

Machine Learning · Computer Science 2024-03-14 Guy Amit , Abigail Goldsteen , Ariel Farkash

Machine Learning (ML) for information security (InfoSec) utilizes distinct data types and formats which require different treatments during optimization/training on raw data. In this paper, we implement a malicious/benign URL predictor…

Cryptography and Security · Computer Science 2020-11-06 Ethan M. Rudd , Ahmed Abdallah

The prompt-based learning paradigm, which bridges the gap between pre-training and fine-tuning, achieves state-of-the-art performance on several NLP tasks, particularly in few-shot settings. Despite being widely applied, prompt-based…

Computation and Language · Computer Science 2024-02-05 Shuai Zhao , Jinming Wen , Luu Anh Tuan , Junbo Zhao , Jie Fu

A backdoor or Trojan attack is an important type of data poisoning attack against deep neural network (DNN) classifiers, wherein the training dataset is poisoned with a small number of samples that each possess the backdoor pattern (usually…

Machine Learning · Computer Science 2023-03-15 H. Wang , S. Karami , O. Dia , H. Ritter , E. Emamjomeh-Zadeh , J. Chen , Z. Xiang , D. J. Miller , G. Kesidis

Deep neural networks (DNNs) have been widely and successfully adopted and deployed in various applications of speech recognition. Recently, a few works revealed that these models are vulnerable to backdoor attacks, where the adversaries can…

Sound · Computer Science 2023-07-18 Hanbo Cai , Pengcheng Zhang , Hai Dong , Yan Xiao , Stefanos Koffas , Yiming Li

The rapid growth of natural language processing (NLP) and pre-trained language models have enabled accurate text classification in a variety of settings. However, text classification models are susceptible to backdoor attacks, where an…

Cryptography and Security · Computer Science 2024-12-30 A. Dilara Yavuz , M. Emre Gursoy

Large language model (LLM) unlearning has become a critical mechanism for removing undesired data, knowledge, or behaviors from pre-trained models while retaining their general utility. Yet, with the rise of open-weight LLMs, we ask: can…

Machine Learning · Computer Science 2025-10-21 Bingqi Shang , Yiwei Chen , Yihua Zhang , Bingquan Shen , Sijia Liu