English
Related papers

Related papers: Test-Time Training Undermines Safety Guardrails

200 papers

Convolutional Neural Networks (ConvNets) are trained offline using the few available data and may therefore suffer from substantial accuracy loss when ported on the field, where unseen input patterns received under unpredictable external…

Computer Vision and Pattern Recognition · Computer Science 2021-05-14 Luca Mocerino , Roberto G. Rizzo , Valentino Peluso , Andrea Calimera , Enrico Macii

Optimizing charging protocols is critical for reducing battery charging time and decelerating battery degradation in applications such as electric vehicles. Recently, reinforcement learning (RL) methods have been adopted for such purposes.…

Systems and Control · Electrical Eng. & Systems 2024-06-19 Myisha A. Chowdhury , Saif S. S. Al-Wahaibi , Qiugang Lu

Can an agent learn efficiently in a noisy and self adapting environment with sequential, non-stationary and non-homogeneous observations? Through trading bots, we illustrate how Deep Reinforcement Learning (DRL) can tackle this challenge.…

Machine Learning · Computer Science 2020-10-19 Eric Benhamou , David Saltiel , Sandrine Ungari , Abhishek Mukhopadhyay , Jamal Atif

The remarkable progress in deep learning (DL) showcases outstanding results in various computer vision tasks. However, adaptation to real-time variations in data distributions remains an important challenge. Test-Time Training (TTT) was…

Computer Vision and Pattern Recognition · Computer Science 2024-11-28 Marco Colussi , Sergio Mascetti , Jose Dolz , Christian Desrosiers

Reinforcement learning (RL) fine-tuning transforms large language models while creating a vulnerability we experimentally verify: Our experiment shows that malicious RL fine-tuning dismantles safety guardrails with remarkable efficiency,…

Machine Learning · Computer Science 2025-05-08 Wenjun Cao

Deep learning models are prone to being fooled by imperceptible perturbations known as adversarial attacks. In this work, we study how equipping models with Test-time Transformation Ensembling (TTE) can work as a reliable defense against…

Machine Learning · Computer Science 2021-07-30 Juan C. Pérez , Motasem Alfarra , Guillaume Jeanneret , Laura Rueda , Ali Thabet , Bernard Ghanem , Pablo Arbeláez

The autonomous AI agents using large language models can create undeniable values in all span of the society but they face security threats from adversaries that warrants immediate protective solutions because trust and safety issues arise.…

Cryptography and Security · Computer Science 2025-06-13 Saikat Barua , Mostafizur Rahman , Md Jafor Sadek , Rafiul Islam , Shehenaz Khaled , Ahmedul Kabir

Recently, safe reinforcement learning (RL) with the actor-critic structure for continuous control tasks has received increasing attention. It is still challenging to learn a near-optimal control policy with safety and convergence…

Machine Learning · Computer Science 2024-02-06 Xinglong Zhang , Yaoqian Peng , Biao Luo , Wei Pan , Xin Xu , Haibin Xie

The primary goal of reinforcement learning is to develop decision-making policies that prioritize optimal performance, frequently without considering safety. In contrast, safe reinforcement learning seeks to reduce or avoid unsafe behavior.…

Machine Learning · Computer Science 2025-06-17 Zahra Shahrooei , Ali Baheri

Single-step adversarial training (AT) has received wide attention as it proved to be both efficient and robust. However, a serious problem of catastrophic overfitting exists, i.e., the robust accuracy against projected gradient descent…

Machine Learning · Computer Science 2022-03-23 Tao Li , Yingwen Wu , Sizhe Chen , Kun Fang , Xiaolin Huang

Pretrained language models often rely on superficial features that appear predictive during training yet fail to generalize at test time, a phenomenon known as shortcut learning. Existing mitigation methods generally operate at training…

Machine Learning · Computer Science 2026-04-15 Jiayi Li , Shijie Tang , Gün Kaynar , Shiyi Du , Carl Kingsford

Retrieval-Augmented Generation (RAG) has emerged as a powerful approach for enhancing large language models' question-answering capabilities through the integration of external knowledge. However, when adapting RAG systems to specialized…

Computation and Language · Computer Science 2026-01-19 Xin Sun , Zhongqi Chen , Qiang Liu , Shu Wu , Bowen Song , Weiqiang Wang , Zilei Wang , Liang Wang

Test-time adaptation (TTA) is an effective approach to mitigate performance degradation of trained models when encountering input distribution shifts at test time. However, existing TTA methods often suffer significant performance drops…

Machine Learning · Computer Science 2025-02-06 Minguk Jang , Hye Won Chung

Adversarial training has been considered an imperative component for safely deploying neural network-based applications to the real world. To achieve stronger robustness, existing methods primarily focus on how to generate strong attacks by…

Machine Learning · Computer Science 2023-09-01 Yeachan Kim , Seongyeon Kim , Ihyeok Seo , Bonggun Shin

Test time adaptation (TTA) equips deep learning models to handle unseen test data that deviates from the training distribution, even when source data is inaccessible. While traditional TTA methods often rely on entropy as a confidence…

Machine Learning · Computer Science 2024-09-17 Afshar Shamsi , Rejisa Becirovic , Ahmadreza Argha , Ehsan Abbasnejad , Hamid Alinejad-Rokny , Arash Mohammadi

Recent progress in AI and Reinforcement learning has shown great success in solving complex problems with high dimensional state spaces. However, most of these successes have been primarily in simulated environments where failure is of…

Artificial Intelligence · Computer Science 2019-03-25 Bharat Prakash , Mohit Khatwani , Nicholas Waytowich , Tinoosh Mohsenin

Adversarial training (AT) is among the most effective techniques to improve model robustness by augmenting training data with adversarial examples. However, most existing AT methods adopt a specific attack to craft adversarial examples,…

Machine Learning · Computer Science 2020-11-20 Yinpeng Dong , Zhijie Deng , Tianyu Pang , Hang Su , Jun Zhu

We propose tttLRM, a novel large 3D reconstruction model that leverages a Test-Time Training (TTT) layer to enable long-context, autoregressive 3D reconstruction with linear computational complexity, further scaling the model's capability.…

Computer Vision and Pattern Recognition · Computer Science 2026-03-03 Chen Wang , Hao Tan , Wang Yifan , Zhiqin Chen , Yuheng Liu , Kalyan Sunkavalli , Sai Bi , Lingjie Liu , Yiwei Hu

In this technical report, we evaluate the adversarial robustness of a very recent method called "Geometry-aware Instance-reweighted Adversarial Training"[7]. GAIRAT reports state-of-the-art results on defenses to adversarial attacks on the…

Machine Learning · Computer Science 2021-03-08 Dorjan Hitaj , Giulio Pagnotta , Iacopo Masi , Luigi V. Mancini

Time Series Classification (TSC) is highly vulnerable to backdoor attacks, posing significant security threats. Existing methods primarily focus on data poisoning during the training phase, designing sophisticated triggers to improve…

Cryptography and Security · Computer Science 2025-02-04 Chang Dong , Zechao Sun , Guangdong Bai , Shuying Piao , Weitong Chen , Wei Emma Zhang