English
Related papers

Related papers: Do Transformers Use their Depth Adaptively? Eviden…

200 papers

We conduct a systematic study of the approximation properties of Transformer for sequence modeling with long, sparse and complicated memory. We investigate the mechanisms through which different components of Transformer, such as the…

Machine Learning · Computer Science 2024-10-31 Mingze Wang , Weinan E

Despite the empirical success of prompt tuning in adapting pretrained language models to new tasks, theoretical analyses of its capabilities remain limited. Existing theoretical work primarily addresses universal approximation properties,…

Machine Learning · Computer Science 2025-09-03 Maxime Meyer , Mario Michelessa , Caroline Chaux , Vincent Y. F. Tan

Language models are increasingly capable, yet still fail at a seemingly simple task of multi-digit multiplication. In this work, we study why, by reverse-engineering a model that successfully learns multiplication via \emph{implicit…

Machine Learning · Computer Science 2025-10-02 Xiaoyan Bai , Itamar Pres , Yuntian Deng , Chenhao Tan , Stuart Shieber , Fernanda Viégas , Martin Wattenberg , Andrew Lee

Existing work on understanding deep learning often employs measures that compress all data-dependent information into a few numbers. In this work, we adopt a perspective based on the role of individual examples. We introduce a measure of…

Machine Learning · Computer Science 2021-06-21 Robert J. N. Baldock , Hartmut Maennel , Behnam Neyshabur

Gradually growing the depth of Transformers during training can not only reduce training cost but also lead to improved reasoning performance, as shown by MIDAS (Saunshi et al., 2024). Thus far, however, a mechanistic understanding of these…

Computation and Language · Computer Science 2025-12-10 Ferdinand Kapl , Emmanouil Angelis , Tobias Höppe , Kaitlin Maile , Johannes von Oswald , Nino Scherrer , Stefan Bauer

Large language models often reason beyond surface tokens, but the internal stage at which token-level information becomes abstract relational structure remains unclear. We investigate this question by analyzing how attention heads and…

Artificial Intelligence · Computer Science 2026-05-22 Junjie Zhang , Zhen Shen , Xisong Dong , Gang Xiong

Deep Learning architectures, and in particular Transformers, are conventionally viewed as a composition of layers. These layers are actually often obtained as the sum of two contributions: a residual path that copies the input and the…

Large language models (LLMs) perform well on multi-hop reasoning, yet how they internally compose multiple facts remains unclear. Recent work proposes \emph{hop-aligned circuit hypothesis}, suggesting that bridge entities are computed…

Computation and Language · Computer Science 2026-01-08 Xukai Liu , Ye Liu , Jipeng Zhang , Yanghai Zhang , Kai Zhang , Qi Liu

Transformer layers, which use an alternating pattern of multi-head attention and multi-layer perceptron (MLP) layers, provide an effective tool for a variety of machine learning problems. As the transformer layers use residual connections…

Machine Learning · Computer Science 2022-12-13 Yaofeng Desmond Zhong , Tongtao Zhang , Amit Chakraborty , Biswadip Dey

Overparameterized transformer networks have obtained state of the art results in various natural language processing tasks, such as machine translation, language modeling, and question answering. These models contain hundreds of millions of…

Machine Learning · Computer Science 2019-09-26 Angela Fan , Edouard Grave , Armand Joulin

Many neural network architectures are known to be Turing Complete, and can thus, in principle implement arbitrary algorithms. However, Transformers are unique in that they can implement gradient-based learning algorithms under simple…

Machine Learning · Computer Science 2024-06-05 Xiang Cheng , Yuxin Chen , Suvrit Sra

We study the approximation capabilities, convergence speeds and on-convergence behaviors of transformers trained on in-context recall tasks -- which requires to recognize the \emph{positional} association between a pair of tokens from…

Machine Learning · Computer Science 2025-10-22 Quan Nguyen , Thanh Nguyen-Tang

Until now, it has been difficult for volumetric super-resolution to utilize the recent advances in transformer-based models seen in 2D super-resolution. The memory required for self-attention in 3D volumes limits the receptive field.…

Computer Vision and Pattern Recognition · Computer Science 2024-12-10 August Leander Høeg , Sophia W. Bardenfleth , Hans Martin Kjer , Tim B. Dyrby , Vedrana Andersen Dahl , Anders Dahl

Transformers achieve state-of-the-art accuracy and robustness across many tasks, but an understanding of their inductive biases and how those biases differ from other neural network architectures remains elusive. In this work, we identify…

Machine Learning · Computer Science 2025-02-14 Bhavya Vasudeva , Deqing Fu , Tianyi Zhou , Elliott Kau , Youqi Huang , Vatsal Sharan

In this paper, we investigate the extrapolation capabilities of implicit deep learning models in handling unobserved data, where traditional deep neural networks may falter. Implicit models, distinguished by their adaptability in layer…

Machine Learning · Computer Science 2024-07-22 Juliette Decugis , Alicia Y. Tsai , Max Emerling , Ashwin Ganesh , Laurent El Ghaoui

In-context learning \ -- performing tasks based on examples given in the prompt \ -- is an important capability that has emerged in large language models and has received significant attention in both theory and practice. Existing…

Machine Learning · Computer Science 2026-05-28 Ruomin Huang , Eshaan Nichani , Jason D. Lee , Rong Ge

Recurrent Neural Networks were, until recently, one of the best ways to capture the timely dependencies in sequences. However, with the introduction of the Transformer, it has been proven that an architecture with only attention-mechanisms…

Machine Learning · Computer Science 2021-08-19 Radostin Cholakov , Todor Kolev

Transfer learning with large pretrained transformer-based language models like BERT has become a dominating approach for most NLP tasks. Simply fine-tuning those large language models on downstream tasks or combining it with task-specific…

Computation and Language · Computer Science 2021-08-06 Wenjuan Han , Bo Pang , Yingnian Wu

Transformer-based Large Language Models (LLMs) have demonstrated powerful in-context learning capabilities. However, their predictions can be disrupted by factually correct context, a phenomenon known as context hijacking, revealing a…

Computation and Language · Computer Science 2025-02-24 Tianle Li , Chenyang Zhang , Xingwu Chen , Yuan Cao , Difan Zou

Transformer architecture has become ubiquitous in the natural language processing field. To interpret the Transformer-based models, their attention patterns have been extensively analyzed. However, the Transformer architecture is not only…

Computation and Language · Computer Science 2021-09-16 Goro Kobayashi , Tatsuki Kuribayashi , Sho Yokoi , Kentaro Inui