English
Related papers

Related papers: Finite-Size Gradient Transport in Large Language M…

200 papers

Large language models (LLMs) have demonstrated emergent in-context learning (ICL) capabilities across a range of tasks, including zero-shot time-series forecasting. We show that text-trained foundation models can accurately extrapolate…

Machine Learning · Computer Science 2026-03-13 Jiajun Bao , Nicolas Boullé , Toni J. B. Liu , Raphaël Sarfati , Christopher J. Earls

Research on Large Language Models (LLMs) increasingly focuses on identifying mechanistic explanations for their behaviors, yet the field lacks clear principles for determining when (and how) findings from one model instance generalize to…

Artificial Intelligence · Computer Science 2025-09-30 Sean Trott

Understanding how latent representations evolve during generation is a central open problem in large language model interpretability. We introduce \textbf{Dynamical Manifold Evolution Theory} (DMET), a phenomenological framework that models…

Computation and Language · Computer Science 2026-05-05 Yukun Zhang , Qi Dong , Mengkang Li

Recent studies show that transformer-based architectures emulate gradient descent during a forward pass, contributing to in-context learning capabilities - an ability where the model adapts to new tasks based on a sequence of prompt…

Statistics Theory · Mathematics 2024-05-13 Karthik Duraisamy

Computation in a typical Transformer-based large language model (LLM) can be characterized by batch size, hidden dimension, number of layers, and sequence length. Until now, system works for accelerating LLM training have focused on the…

Machine Learning · Computer Science 2023-10-05 Sam Ade Jacobs , Masahiro Tanaka , Chengming Zhang , Minjia Zhang , Shuaiwen Leon Song , Samyam Rajbhandari , Yuxiong He

Large-scale distributed training of deep neural networks results in models with worse generalization performance as a result of the increase in the effective mini-batch size. Previous approaches attempt to address this problem by varying…

Machine Learning · Computer Science 2020-02-17 Kazuki Osawa , Yohei Tsuji , Yuichiro Ueno , Akira Naruse , Chuan-Sheng Foo , Rio Yokota

Aligned large language models (LLMs) demonstrate exceptional capabilities in task-solving, following instructions, and ensuring safety. However, the continual learning aspect of these aligned LLMs has been largely overlooked. Existing…

Computation and Language · Computer Science 2023-10-11 Xiao Wang , Yuansen Zhang , Tianze Chen , Songyang Gao , Senjie Jin , Xianjun Yang , Zhiheng Xi , Rui Zheng , Yicheng Zou , Tao Gui , Qi Zhang , Xuanjing Huang

Understanding why Transformers perform so well remains challenging due to their non-convex optimization landscape. In this work, we analyze a shallow Transformer with $m$ independent heads trained by projected gradient descent in the kernel…

Machine Learning · Computer Science 2026-04-03 Enes Arda , Semih Cayci , Atilla Eryilmaz

We investigate whether independently trained language models converge to geometrically compatible latent representations, and whether this compatibility can be exploited to correct model behavior at inference time without any weight…

Machine Learning · Computer Science 2026-03-24 Marcus Armstrong , Navid Ayoobi , Arjun Mukherjee

Language models (LMs) require robust episodic grounding-the capacity to learn from and apply past experiences-to excel at physical planning tasks. Current episodic grounding approaches struggle with scalability and integration, limiting…

Computation and Language · Computer Science 2025-06-03 Chunhui Zhang , Sirui , Wang , Zhongyu Ouyang , Xiangchi Yuan , Soroush Vosoughi

Uncovering early-stage metrics that reflect final model performance is one core principle for large-scale pretraining. The existing scaling law demonstrates the power-law correlation between pretraining loss and training flops, which serves…

Gradient descent (GD) on logistic regression has many fascinating properties. When the dataset is linearly separable, it is known that the iterates converge in direction to the maximum-margin separator regardless of how large the step size…

Machine Learning · Computer Science 2025-07-16 Si Yi Meng , Baptiste Goujaud , Antonio Orvieto , Christopher De Sa

To fully unlock the benefits of multiple-input multiple-output (MIMO) networks, downlink channel state information (CSI) is required at the base station (BS). In frequency division duplex (FDD) systems, the CSI is acquired through a…

Information Theory · Computer Science 2023-05-23 Matteo Nerini , Valentina Rizzello , Michael Joham , Wolfgang Utschick , Bruno Clerckx

Recent advances in large language models (LLMs) have intensified the need to deliver both rapid responses and high-quality outputs. More powerful models yield better results but incur higher inference latency, whereas smaller models are…

Distributed, Parallel, and Cluster Computing · Computer Science 2025-10-01 Youhe Jiang , Fangcheng Fu , Wanru Zhao , Stephan Rabanser , Jintao Zhang , Nicholas D. Lane , Binhang Yuan

We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When we train increasingly large neural networks from-scratch on a fixed-size dataset, they eventually become data-limited…

Machine Learning · Computer Science 2021-02-03 Danny Hernandez , Jared Kaplan , Tom Henighan , Sam McCandlish

Recent advances in large language models have demonstrated the effectiveness of length scaling during post-training, yet its potential in pre-training remains underexplored. We present the Parallel Hidden Decoding Transformer…

Computation and Language · Computer Science 2025-04-25 Bohong Wu , Shen Yan , Sijun Zhang , Jianqiao Lu , Yutao Zeng , Ya Wang , Xun Zhou

We investigate a scalar partial differential equation model for the formation of biological transportation networks. Starting from a discrete graph-based formulation on equilateral triangulations, we rigorously derive the corresponding…

Analysis of PDEs · Mathematics 2025-10-20 Jan Haskovec , Peter Markowich , Stefano Zampini

Data-parallel (DP) training with synchronous all-reduce is a dominant paradigm for full-parameter fine-tuning of large language models (LLMs). While parameter synchronization guarantees numerical equivalence of model weights after each…

Machine Learning · Computer Science 2026-02-25 Hong Li , Zhen Zhou , Honggang Zhang , Yuping Luo , Xinyue Wang , Han Gong , Zhiyuan Liu

Gradient descent and stochastic gradient descent are central to modern machine learning, yet their behavior under large step sizes remains theoretically unclear. Recent work suggests that acceleration often arises near the edge of…

Machine Learning · Computer Science 2026-03-02 Sacchit Kale , Piyushi Manupriya , Pierre Marion , Francis Bach , Anant Raj

As Large Language Models (LLMs) grow dramatically in size, there is an increasing trend in compressing and speeding up these models. Previous studies have highlighted the usefulness of gradients for importance scoring in neural network…

Computation and Language · Computer Science 2024-07-17 Hongrong Cheng , Miao Zhang , Javen Qinfeng Shi