English
Related papers

Related papers: The Geometry of Alignment Collapse: When Fine-Tuni…

200 papers

Accurate uncertainty quantification is crucial for the safe deployment of machine learning models, and prior research has demonstrated improvements in the calibration of modern language models (LMs). We study in-context learning (ICL), a…

Computation and Language · Computer Science 2024-03-29 Hanlin Zhang , Yi-Fan Zhang , Yaodong Yu , Dhruv Madeka , Dean Foster , Eric Xing , Himabindu Lakkaraju , Sham Kakade

Machine learning has demonstrated remarkable prediction accuracy over i.i.d data, but the accuracy often drops when tested with data from another distribution. In this paper, we aim to offer another view of this problem in a perspective…

Machine Learning · Computer Science 2022-06-20 Haohan Wang , Zeyi Huang , Hanlin Zhang , Yong Jae Lee , Eric Xing

Large language models exhibit remarkable performance across diverse tasks through pre-training and fine-tuning paradigms. However, continual fine-tuning on sequential tasks induces catastrophic forgetting, where newly acquired knowledge…

Machine Learning · Computer Science 2026-01-27 Olaf Yunus Laitinen Imanov

Large learning rates, when applied to gradient descent for nonconvex optimization, yield various implicit biases including the edge of stability (Cohen et al., 2021), balancing (Wang et al., 2022), and catapult (Lewkowycz et al., 2020).…

Machine Learning · Computer Science 2023-12-13 Yuqing Wang , Zhenghao Xu , Tuo Zhao , Molei Tao

The highly non-linear nature of deep neural networks causes them to be susceptible to adversarial examples and have unstable gradients which hinders interpretability. However, existing methods to solve these issues, such as adversarial…

Machine Learning · Computer Science 2023-01-11 Suraj Srinivas , Kyle Matoba , Himabindu Lakkaraju , Francois Fleuret

Large language models (LLMs) are increasingly deployed in security-sensitive applications, where they must follow system- or developer-specified instructions that define the intended task behavior, while completing benign user requests.…

Cryptography and Security · Computer Science 2026-01-13 Shawn Li , Chenxiao Yu , Zhiyu Ni , Hao Li , Charith Peris , Chaowei Xiao , Yue Zhao

Alignment faking is a form of strategic deception in AI in which models selectively comply with training objectives when they infer that they are in training, while preserving different behavior outside training. The phenomenon was first…

The fine-tuning paradigm has emerged as a prominent approach for addressing long-tail learning tasks in the era of foundation models. However, the impact of fine-tuning strategies on long-tail learning performance remains unexplored. In…

Computer Vision and Pattern Recognition · Computer Science 2025-04-21 Jiang-Xin Shi , Tong Wei , Yu-Feng Li

AI alignment research aims to develop techniques to ensure that AI systems do not cause harm. However, every alignment technique has failure modes, which are conditions in which there is a non-negligible chance that the technique fails to…

Artificial Intelligence · Computer Science 2025-10-14 Leonard Dung , Florian Mai

Training large language models is an expensive, compute-bound process that must be repeated as models scale, algorithms improve, and new data is collected. To address this, next-generation hardware accelerators increasingly support…

Machine Learning · Computer Science 2025-06-27 Huangyuan Su , Mujin Kwun , Stephanie Gil , Sham Kakade , Nikhil Anand

Recent defenses for safeguarding open-weight large language models (LLMs) are intended to prevent adversarial usage. Underlying these defenses is an assumption that new harmful behavior is learned through fine-tuning rather than elicited by…

Machine Learning · Computer Science 2026-05-27 Kevin Kuo , Chhavi Yadav , Virginia Smith

Vision-language models encode continuous geometry that their text pathway fails to express: a 6,000-parameter linear probe extracts hand joint angles at 6.1 degrees MAE from frozen features, while the best text output achieves only 20.0…

Computer Vision and Pattern Recognition · Computer Science 2026-03-09 Yakov Pyotr Shkolnikov

Machine unlearning, an emerging research topic focusing on compliance with data privacy regulations, enables trained models to remove the information learned from specific data. While many existing methods indirectly address this issue by…

Machine Learning · Computer Science 2024-12-24 Seonguk Seo , Dongwan Kim , Bohyung Han

A key concern with the concept of "alignment" is the implicit question of "alignment to what?". AI systems are increasingly used across the world, yet safety alignment is often focused on homogeneous monolingual settings. Additionally,…

Computation and Language · Computer Science 2024-07-09 Aakanksha , Arash Ahmadian , Beyza Ermis , Seraphina Goldfarb-Tarrant , Julia Kreutzer , Marzieh Fadaee , Sara Hooker

Vision-Language Models (VLMs), such as CLIP, have achieved significant zero-shot performance on downstream tasks with various fine-tuning adaptation methods. However, recent studies have proven that adversarial attacks can significantly…

Computer Vision and Pattern Recognition · Computer Science 2026-05-20 Jia-Wei Hai , Yijun Wang , Xiu-Shen Wei

As advancements in large language models (LLMs) continue and the demand for personalized models increases, parameter-efficient fine-tuning (PEFT) methods (e.g., LoRA) will become essential due to their efficiency in reducing computation…

Machine Learning · Computer Science 2025-01-06 Mingjie Li , Wai Man Si , Michael Backes , Yang Zhang , Yisen Wang

A widely believed explanation for the remarkable generalization capacities of overparameterized neural networks is that the optimization algorithms used for training induce an implicit bias towards benign solutions. To grasp this…

Machine Learning · Computer Science 2025-12-19 Maria Matveev , Vit Fojtik , Hung-Hsu Chou , Gitta Kutyniok , Johannes Maly

This work introduces a novel framework for evaluating LLMs' capacity to balance instruction-following with critical reasoning when presented with multiple-choice questions containing no valid answers. Through systematic evaluation across…

Computation and Language · Computer Science 2025-06-03 Gracjan Góral , Emilia Wiśnios , Piotr Sankowski , Paweł Budzianowski

Loss of trainability refers to a phenomenon in continual learning where parameter updates no longer make progress on the optimization objective, so accuracy stalls or degrades as the learning problem changes over time. In this paper, we…

Machine Learning · Computer Science 2025-12-11 Gunbir Singh Baveja , Alex Lewandowski , Mark Schmidt

While deep learning has significantly advanced accident anticipation, the robustness of these safety-critical systems against real-world perturbations remains a major challenge. We reveal that state-of-the-art models like CRASH, despite…

Machine Learning · Computer Science 2026-04-03 Wenjing Wang , Wenxuan Wang , Songning Lai
‹ Prev 1 8 9 10 Next ›