中文
相关论文

相关论文: Grokking phase transitions in learning local rules…

200 篇论文

Training dynamics during grokking concentrate along a small number of dominant update directions -- the spectral edge -- which reliably distinguishes grokking from non-grokking regimes. We show that standard mechanistic interpretability…

机器学习 · 计算机科学 2026-04-09 Yongzhong Xu

Gradual argumentation frameworks represent arguments and their relationships in a weighted graph. Their graphical structure and intuitive semantics makes them a potentially interesting tool for interpretable machine learning. It has been…

机器学习 · 计算机科学 2021-06-28 Jonathan Spieler , Nico Potyka , Steffen Staab

We develop a mathematical model to describe the persistence of rule-breaking behaviors in societies, such as traffic violations, disregard for legal restrictions and other forms of noncompliance. Using a replicator-type dynamics with…

物理与社会 · 物理学 2026-05-27 Nuno Crokidakis

This paper considers a class of probabilistic cellular automata undergoing a phase transition with an absorbing state. Denoting by ${\mathcal{U}}(x)$ the neighbourhood of site $x$, the transition probability is $T(\eta_x = 1 |…

数学物理 · 物理学 2015-05-19 Lorenzo Taggi

Motivated by empirical observations of prolonged plateaus and stage-wise progression during training, we investigate the loss landscape of transformer models trained on in-context next-token prediction tasks. In particular, we focus on…

机器学习 · 计算机科学 2025-08-20 Aditya Varre , Gizem Yüce , Nicolas Flammarion

The `random intersection graph with communities' models networks with communities, assuming an underlying bipartite structure of groups and individuals. Each group has its own internal structure described by a (small) graph, while groups…

概率论 · 数学 2019-10-23 Remco van der Hofstad , Júlia Komjáthy , Viktória Vadon

We study the generalization ability of a simple perceptron which learns unlearnable rules. The rules are presented by a teacher perceptron with a non-monotonic transfer function. The student is trained in the on-line mode. The asymptotic…

凝聚态物理 · 物理学 2009-10-30 Jun-ichi Inoue , Hidetoshi Nishimori , Yoshiyuki Kabashima

The problem of learning tree-structured Gaussian graphical models from independent and identically distributed (i.i.d.) samples is considered. The influence of the tree structure and the parameters of the Gaussian distribution on the…

机器学习 · 统计学 2010-05-06 Vincent Y. F. Tan , Animashree Anandkumar , Alan S. Willsky

Training Neural Networks (NNs) without overfitting is difficult; detecting that overfitting is difficult as well. We present a novel Random Matrix Theory method that detects the onset of overfitting in deep learning models without access to…

机器学习 · 计算机科学 2026-05-15 Hari K. Prakash , Charles H Martin

Learning from Demonstration depends on a robot learner generalising its learned model to unseen conditions, as it is not feasible for a person to provide a demonstration set that accounts for all possible variations in non-trivial tasks.…

机器人学 · 计算机科学 2019-03-05 Aran Sena , Brendan Michael , Matthew Howard

Graph convolutional networks (GCNs) have recently achieved great empirical success in learning graph-structured data. To address its scalability issue due to the recursive embedding of neighboring features, graph topology sampling has been…

机器学习 · 计算机科学 2023-12-12 Hongkang Li , Meng Wang , Sijia Liu , Pin-Yu Chen , Jinjun Xiong

In the Big Data era, with the ubiquity of geolocation sensors in particular, massive datasets exhibiting a possibly complex spatial dependence structure are becoming increasingly available. In this context, the standard probabilistic theory…

机器学习 · 统计学 2024-02-05 Emilia Siviero , Emilie Chautru , Stephan Clémençon

We review the application of Statistical Mechanics methods to the study of online learning of a drifting concept in the limit of large systems. The model where a feed-forward network learns from examples generated by a time dependent…

无序系统与神经网络 · 物理学 2007-05-23 Renato Vicente , Osame Kinouchi , Nestor Caticha

We propose a novel local learning rule for spiking neural networks in which spike propagation times undergo activity-dependent plasticity. Our plasticity rule aligns pre-synaptic spike times to produce a stronger and more rapid response.…

神经与进化计算 · 计算机科学 2022-11-16 Jørgen Jensen Farner , Ola Huse Ramstad , Stefano Nichele , Kristine Heiney

The state of many physical, biological and socio-technical systems evolves by combining smooth local transitions and abrupt resetting events to a set of reference values. The inclusion of the resetting mechanism not only provides the…

统计力学 · 物理学 2022-12-21 Oriol Artime

Graph Neural Networks (GNNs) rely on graph convolutions to exploit meaningful patterns in networked data. Based on matrix multiplications, convolutions incur in high computational costs leading to scalability limitations in practice. To…

机器学习 · 计算机科学 2022-10-28 Juan Cervino , Luana Ruiz , Alejandro Ribeiro

Transformers have proven highly effective across various applications, especially in handling sequential data such as natural languages and time series. However, transformer models often lack clear interpretability, and the success of…

机器学习 · 计算机科学 2025-12-01 Wei Shi , Yuan Cao

We introduce a new class of probabilistic cellular automata that are capable of exhibiting rich dynamics such as synchronization and ergodicity and can be easily inferred from data. The system is a finite-state locally interacting Markov…

概率论 · 数学 2025-05-23 Erhan Bayraktar , Fei Lu , Mauro Maggioni , Ruoyu Wu , Sichen Yang

Recent non-asymptotic analyses have substantially advanced the theory of distributional policy evaluation, but they largely concern synchronous full-state updates under a generative model, model-based estimators, accelerated variants, or…

机器学习 · 计算机科学 2026-05-11 Ege C. Kaya , Abolfazl Hashemi

Transformers serve as the foundational architecture for many successful large-scale models, demonstrating the ability to overfit the training data while maintaining strong generalization on unseen data, a phenomenon known as benign…

机器学习 · 计算机科学 2025-02-19 Yingying Zhang , Zhenyu Wu , Jian Li , Yong Liu