中文
相关论文

相关论文: Information Bottleneck Revisited: Posterior Probab…

200 篇论文

The Information Bottleneck (IB) objective uses information theory to formulate a task-performance versus robustness trade-off. It has been successfully applied in the standard discriminative classification setting. We pose the question…

机器学习 · 计算机科学 2021-01-13 Lynton Ardizzone , Radek Mackowiak , Carsten Rother , Ullrich Köthe

Decisions of complex language understanding models can be rationalized by limiting their inputs to a relevant subsequence of the original text. A rationale should be as concise as possible without significantly degrading task performance,…

计算与语言 · 计算机科学 2020-11-04 Bhargavi Paranjape , Mandar Joshi , John Thickstun , Hannaneh Hajishirzi , Luke Zettlemoyer

The Information Bottleneck (IB) principle facilitates effective representation learning by preserving label-relevant information while compressing irrelevant information. However, its strong reliance on accurate labels makes it inherently…

机器学习 · 计算机科学 2025-12-12 Yi Huang , Qingyun Sun , Yisen Gao , Haonan Yuan , Xingcheng Fu , Jianxin Li

In many applications, it is desirable to extract only the relevant information from complex input data, which involves making a decision about which input features are relevant. The information bottleneck method formalizes this as an…

机器学习 · 统计学 2020-04-28 Anirudh Goyal , Yoshua Bengio , Matthew Botvinick , Sergey Levine

Avoiding overfitting is a central challenge in machine learning, yet many large neural networks readily achieve zero training loss. This puzzling contradiction necessitates new approaches to the study of overfitting. Here we quantify…

信息论 · 计算机科学 2022-10-13 Vudtiwat Ngampruetikorn , David J. Schwab

The partial information decomposition (PID) aims to quantify the amount of redundant information that a set of sources provides about a target. Here, we show that this goal can be formulated as a type of information bottleneck (IB) problem,…

信息论 · 计算机科学 2024-06-28 Artemy Kolchinsky

We show that if the conditional distribution p(C | T) factors through a sufficient statistic {\phi}(T), then the Information Bottleneck (IB) problem for (T, C) is exactly equivalent to the IB problem for ({\phi}(T), C). The reduction is…

信息论 · 计算机科学 2026-04-30 Joss Armstrong

The Information Bottleneck principle offers both a mechanism to explain how deep neural networks train and generalize, as well as a regularized objective with which to train models. However, multiple competing objectives are proposed in the…

机器学习 · 计算机科学 2021-01-06 Andreas Kirsch , Clare Lyle , Yarin Gal

This paper investigates task-oriented communication for edge inference, where a low-end edge device transmits the extracted feature vector of a local data sample to a powerful edge server for processing. It is critical to encode the data…

信号处理 · 电气工程与系统科学 2023-01-19 Jiawei Shao , Yuyi Mao , Jun Zhang

This tutorial paper focuses on the variants of the bottleneck problem taking an information theoretic perspective and discusses practical methods to solve it, as well as its connection to coding and learning aspects. The intimate…

信息论 · 计算机科学 2020-02-19 Abdellatif Zaidi , Inaki Estella Aguerri , Shlomo Shamai

In many complex systems, we observe that `interesting behaviour' is often the consequence of a system exploiting the existence of an Information Bottleneck (IB). These bottlenecks can occur at different scales, between individuals or…

物理与社会 · 物理学 2023-08-02 Michael Crosscombe , Hiroki Sato

The training dynamics of hidden layers in deep learning are poorly understood in theory. Recently, the Information Plane (IP) was proposed to analyze them, which is based on the information-theoretic concept of mutual information (MI). The…

机器学习 · 计算机科学 2020-10-06 Nicolás I. Tapia , Pablo A. Estévez

Deep learning representations are often difficult to interpret, which can hinder their deployment in sensitive applications. Concept Bottleneck Models (CBMs) have emerged as a promising approach to mitigate this issue by learning…

机器学习 · 计算机科学 2026-01-30 Antonio Almudévar , José Miguel Hernández-Lobato , Alfonso Ortega

The information bottleneck principle provides an information-theoretic method for representation learning, by training an encoder to retain all information which is relevant for predicting the label while minimizing the amount of other,…

机器学习 · 计算机科学 2020-02-19 Marco Federici , Anjan Dutta , Patrick Forré , Nate Kushman , Zeynep Akata

When presented with a data stream of two statistically dependent variables, predicting the future of one of the variables (the target stream) can benefit from information about both its history and the history of the other variable (the…

机器学习 · 计算机科学 2023-03-10 Damjan Kalajdzievski , Ximeng Mao , Pascal Fortier-Poisson , Guillaume Lajoie , Blake Richards

Leveraging high-quality joint representations from multimodal data can greatly enhance model performance in various machine-learning based applications. Recent multimodal learning methods, based on the multimodal information bottleneck…

机器学习 · 计算机科学 2025-05-27 Qilong Wu , Yiyang Shao , Jun Wang , Xiaobo Sun

In this paper, we consider establishing a formal connection between two distinct tree-abstraction problems inspired by the information-bottleneck (IB) method. Specifically, we consider the hard- and soft-constrained formulations that have…

信息论 · 计算机科学 2025-12-02 Daniel T. Larsson , Dipankar Maity , Panagiotis Tsiotras

We present a variational approximation to the information bottleneck of Tishby et al. (1999). This variational approach allows us to parameterize the information bottleneck model using a neural network and leverage the reparameterization…

机器学习 · 计算机科学 2019-10-25 Alexander A. Alemi , Ian Fischer , Joshua V. Dillon , Kevin Murphy

Based on the notion of information bottleneck (IB), we formulate a quantization problem called "IB quantization". We show that IB quantization is equivalent to learning based on the IB principle. Under this equivalence, the standard neural…

机器学习 · 计算机科学 2019-02-13 Hongyu Guo , Yongyi Mao , Ali Al-Bashabsheh , Richong Zhang

Integration of data from multiple omics techniques is becoming increasingly important in biomedical research. Due to non-uniformity and technical limitations in omics platforms, such integrative analyses on multiple omics, which we refer to…

机器学习 · 计算机科学 2021-02-11 Changhee Lee , Mihaela van der Schaar