中文
相关论文

相关论文: A Theoretical Analysis on Feature Learning in Neur…

200 篇论文

Neural networks outperform kernel methods, sometimes by orders of magnitude, e.g. on staircase functions. This advantage stems from the ability of neural networks to learn features, adapting their hidden representations to better capture…

In recent years neural networks have achieved impressive results on many technological and scientific tasks. Yet, the mechanism through which these models automatically select features, or patterns in data, for prediction remains unclear.…

机器学习 · 计算机科学 2023-05-11 Adityanarayanan Radhakrishnan , Daniel Beaglehole , Parthe Pandit , Mikhail Belkin

Deep learning has arguably achieved tremendous success in recent years. In simple words, deep learning uses the composition of many nonlinear functions to model the complex dependency between input features and labels. While neural networks…

机器学习 · 统计学 2019-04-16 Jianqing Fan , Cong Ma , Yiqiao Zhong

Neural networks have achieved remarkable empirical performance, while the current theoretical analysis is not adequate for understanding their success, e.g., the Neural Tangent Kernel approach fails to capture their key feature learning…

机器学习 · 计算机科学 2023-10-20 Zhenmei Shi , Junyi Wei , Yingyu Liang

Feature learning is thought to be one of the fundamental reasons for the success of deep neural networks. It is rigorously known that in two-layer fully-connected neural networks under certain conditions, one step of gradient descent on the…

机器学习 · 统计学 2025-04-11 Behrad Moniri , Donghwan Lee , Hamed Hassani , Edgar Dobriban

One of the central questions in the theory of deep learning is to understand how neural networks learn hierarchical features. The ability of deep networks to extract salient features is crucial to both their outstanding generalization…

机器学习 · 计算机科学 2025-04-03 Eshaan Nichani , Alex Damian , Jason D. Lee

Neural networks posses the crucial ability to generate meaningful representations of task-dependent features. Indeed, with appropriate scaling, supervised learning in neural networks can result in strong, task-dependent feature learning.…

机器学习 · 计算机科学 2025-05-13 Alexander van Meegen , Haim Sompolinsky

Exploiting data invariances is crucial for efficient learning in both artificial and biological neural circuits. Understanding how neural networks can discover appropriate representations capable of harnessing the underlying symmetries of…

无序系统与神经网络 · 物理学 2022-10-17 Alessandro Ingrosso , Sebastian Goldt

Often in language and other areas of cognition, whether two components of an object are identical or not determines if it is well formed. We call such constraints identity effects. When developing a system to learn well-formedness from…

机器学习 · 计算机科学 2022-03-03 S. Brugiapaglia , M. Liu , P. Tupper

The ability of learning useful features is one of the major advantages of neural networks. Although recent works show that neural network can operate in a neural tangent kernel (NTK) regime that does not allow feature learning, many works…

机器学习 · 计算机科学 2024-11-06 Mo Zhou , Rong Ge

Feature Learning aims to extract relevant information contained in data sets in an automated fashion. It is driving force behind the current deep learning trend, a set of methods that have had widespread empirical success. What is lacking…

机器学习 · 统计学 2015-04-02 Brendan van Rooyen , Robert C. Williamson

Recently, a spate of papers have provided positive theoretical results for training over-parameterized neural networks (where the network size is larger than what is needed to achieve low error). The key insight is that with sufficient…

机器学习 · 计算机科学 2022-03-01 Gilad Yehudai , Ohad Shamir

According to a popular viewpoint, neural networks learn from data by first identifying low-dimensional representations, and subsequently fitting the best model in this space. Recent works provide a formalization of this phenomenon when…

机器学习 · 计算机科学 2026-02-27 Andrea Montanari , Zihao Wang

We address the challenging problem of deep representation learning--the efficient adaption of a pre-trained deep network to different tasks. Specifically, we propose to explore gradient-based features. These features are gradients of the…

机器学习 · 计算机科学 2020-04-14 Fangzhou Mu , Yingyu Liang , Yin Li

Significant theoretical work has established that in specific regimes, neural networks trained by gradient descent behave like kernel methods. However, in practice, it is known that neural networks strongly outperform their associated…

机器学习 · 计算机科学 2022-07-01 Alex Damian , Jason D. Lee , Mahdi Soltanolkotabi

The predominant success of diffusion models in generative modeling has spurred significant interest in understanding their theoretical foundations. In this work, we propose a feature learning framework aimed at analyzing and comparing the…

机器学习 · 统计学 2025-03-04 Andi Han , Wei Huang , Yuan Cao , Difan Zou

Recent success in training deep neural networks have prompted active investigation into the features learned on their intermediate layers. Such research is difficult because it requires making sense of non-linear computations performed by…

机器学习 · 计算机科学 2016-03-01 Yixuan Li , Jason Yosinski , Jeff Clune , Hod Lipson , John Hopcroft

Often in language and other areas of cognition, whether two components of an object are identical or not determine whether it is well formed. We call such constraints identity effects. When developing a system to learn well-formedness from…

计算与语言 · 计算机科学 2020-05-12 Simone Brugiapaglia , Matthew Liu , Paul Tupper

Feature learning is widely regarded as the key mechanism distinguishing neural networks from fixed-kernel methods, yet its impact on the induced function space remains poorly understood. In this work, we precisely characterize how the…

机器学习 · 统计学 2026-05-19 João Lobo , Bruno Loureiro , Long Tran-Than , Fanghui Liu

With the emergence of large-scale pre-trained neural networks, methods to adapt such "foundation" models to data-limited downstream tasks have become a necessity. Fine-tuning, preference optimization, and transfer learning have all been…

机器学习 · 统计学 2025-07-09 Javan Tahir , Surya Ganguli , Grant M. Rotskoff
‹ 上一页 1 2 3 10 下一页 ›