Imagen de Portada

Which Frequencies do CNNs Need? Emergent Bottleneck Structure in Feature Learning

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Wen, Yuxiao, Jacot, Arthur
Formato:	Preprint
Publicado:	2024
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Bottleneck Structure in Learned Features: Low-Dimension vs Regularity Tradeoff
por: Jacot, Arthur
Publicado: (2023)

Hamiltonian Mechanics of Feature Learning: Bottleneck Structure in Leaky ResNets
por: Jacot, Arthur, et al.
Publicado: (2024)

How DNNs break the Curse of Dimensionality: Compositionality and Symmetry Learning
por: Jacot, Arthur, et al.
Publicado: (2024)

Mixed Dynamics In Linear Networks: Unifying the Lazy and Active Regimes
por: Tu, Zhenfeng, et al.
Publicado: (2024)

Saddle-To-Saddle Dynamics in Deep ReLU Networks: Low-Rank Bias in the First Saddle Escape
por: Bantzis, Ioannis, et al.
Publicado: (2025)

Graph Structure Learning with Temporal Graph Information Bottleneck for Inductive Representation Learning
por: Xiong, Jiafeng, et al.
Publicado: (2025)

Learning to Intervene on Concept Bottlenecks
por: Steinmann, David, et al.
Publicado: (2023)

Group and Exclusive Sparse Regularization-based Continual Learning of CNNs
por: Tousside, Basile, et al.
Publicado: (2026)

Understanding Emergent Abilities of Language Models from the Loss Perspective
por: Du, Zhengxiao, et al.
Publicado: (2024)

These Are Not All the Features You Are Looking For: A Fundamental Bottleneck in Supervised Pretraining
por: Yang, Xingyu Alice, et al.
Publicado: (2025)

GaGSL: Global-augmented Graph Structure Learning via Graph Information Bottleneck
por: Li, Shuangjie, et al.
Publicado: (2024)

Understanding Emergent Misalignment via Feature Superposition Geometry
por: Minegishi, Gouki, et al.
Publicado: (2026)

The Geometry of Concepts: Sparse Autoencoder Feature Structure
por: Li, Yuxiao, et al.
Publicado: (2024)

Return-to-Go Is More Than a Number: Q-Guided Alignment for Return-Conditioned Supervised Learning
por: Yang, Yuxiao, et al.
Publicado: (2026)

On The Potential of The Fractal Geometry and The CNNs Ability to Encode it
por: Zini, Julia El, et al.
Publicado: (2024)

Thought Anchors: Which LLM Reasoning Steps Matter?
por: Bogdan, Paul C., et al.
Publicado: (2025)

Deep Learning with CNNs: A Compact Holistic Tutorial with Focus on Supervised Regression (Preprint)
por: Tejeda, Yansel Gonzalez, et al.
Publicado: (2024)

Credal Concept Bottleneck Models: Structural Separation of Epistemic and Aleatoric Uncertainty
por: Mukherjee, Tanmoy, et al.
Publicado: (2026)

Is Value Learning Really the Main Bottleneck in Offline RL?
por: Park, Seohong, et al.
Publicado: (2024)

Learning Concept Bottleneck Models from Mechanistic Explanations
por: De Santis, Antonio, et al.
Publicado: (2026)

IBNorm: Information-Bottleneck Inspired Normalization for Representation Learning
por: Zou, Xiandong, et al.
Publicado: (2025)

Prediction Bottlenecks Don't Discover Causal Structure (But Here's What They Actually Do)
por: Lade, Ankit Hemant, et al.
Publicado: (2026)

TimeX++: Learning Time-Series Explanations with Information Bottleneck
por: Liu, Zichuan, et al.
Publicado: (2024)

Learning Fair Graph Representations with Multi-view Information Bottleneck
por: Liu, Chuxun, et al.
Publicado: (2025)

Phase-Aware Deep Learning with Complex-Valued CNNs for Audio Signal Applications
por: Agrawal, Naman
Publicado: (2025)

The Interaction Bottleneck of Deep Neural Networks: Discovery, Proof, and Modulation
por: Deng, Huiqi, et al.
Publicado: (2025)

Handling Long and Richly Constrained Tasks through Constrained Hierarchical Reinforcement Learning
por: Lu, Yuxiao, et al.
Publicado: (2023)

No More Adam: Learning Rate Scaling at Initialization is All You Need
por: Xu, Minghao, et al.
Publicado: (2024)

Deep Learning as a Convex Paradigm of Computation: Minimizing Circuit Size with ResNets
por: Jacot, Arthur
Publicado: (2025)

Shapley Neuron Values for Continual Learning: Which Neurons Matter Most?
por: Vahedifar, Mohammad Ali, et al.
Publicado: (2026)

SL-CBM: Enhancing Concept Bottleneck Models with Semantic Locality for Better Interpretability
por: Zhang, Hanwei, et al.
Publicado: (2026)

Delayed Bottlenecking: Alleviating Forgetting in Pre-trained Graph Neural Networks
por: Zhao, Zhe, et al.
Publicado: (2024)

Invariant Graph Learning Meets Information Bottleneck for Out-of-Distribution Generalization
por: Mao, Wenyu, et al.
Publicado: (2024)

Dynamic Graph Information Bottleneck
por: Yuan, Haonan, et al.
Publicado: (2024)

Counterfactual Concept Bottleneck Models
por: Dominici, Gabriele, et al.
Publicado: (2024)

Object Centric Concept Bottlenecks
por: Steinmann, David, et al.
Publicado: (2025)

Mixture of Concept Bottleneck Experts
por: De Santis, Francesco, et al.
Publicado: (2026)

Conditional Clifford-Steerable CNNs with Complete Kernel Basis for PDE Modeling
por: Szarvas, Bálint László, et al.
Publicado: (2025)

From Predictive Importance to Causality: Which Machine Learning Model Reflects Reality?
por: Arshad, Muhammad Arbab, et al.
Publicado: (2024)

Policy Learning with a Language Bottleneck
por: Srivastava, Megha, et al.
Publicado: (2024)