English
Related papers

Related papers: Transformers Generalize Linearly

200 papers

While recurrent models have been effective in NLP tasks, their performance on context-free languages (CFLs) has been found to be quite weak. Given that CFLs are believed to capture important phenomena such as hierarchical structure in…

Computation and Language · Computer Science 2020-11-10 Satwik Bhattamishra , Kabir Ahuja , Navin Goyal

Linguists have long held that a key aspect of natural language syntax is the recursive organization of language units into constituent structures, and research has suggested that current state-of-the-art language models lack an inherent…

Computation and Language · Computer Science 2024-11-27 Michael Ginn

That shared features between train and test data are required for generalisation in artificial neural networks has been a common assumption of both proponents and critics of these models. Here, we show that convolutional architectures avoid…

Neural and Evolutionary Computing · Computer Science 2021-07-15 Jeff Mitchell , Jeffrey S. Bowers

This paper investigates the structure of linear operators introduced in Hernandez et al. [2023] that decode specific relational facts in transformer language models. We extend their single-relation findings to a collection of relations and…

Computation and Language · Computer Science 2025-10-31 Miranda Anna Christ , Adrián Csiszárik , Gergely Becsó , Dániel Varga

Transformer language models have received widespread public attention, yet their generated text is often surprising even to NLP researchers. In this survey, we discuss over 250 recent studies of English language model behavior before…

Computation and Language · Computer Science 2023-08-29 Tyler A. Chang , Benjamin K. Bergen

The architecture of Transformer is based entirely on self-attention, and has been shown to outperform models that employ recurrence on sequence transduction tasks such as machine translation. The superior performance of Transformer has been…

Machine Learning · Computer Science 2019-08-29 Vikas K. Garg , Inderjit S. Dhillon , Hsiang-Fu Yu

The ability to extrapolate from short problem instances to longer ones is an important form of out-of-distribution generalization in reasoning tasks, and is crucial when learning from datasets where longer problem instances are rare. These…

Computation and Language · Computer Science 2022-11-15 Cem Anil , Yuhuai Wu , Anders Andreassen , Aitor Lewkowycz , Vedant Misra , Vinay Ramasesh , Ambrose Slone , Guy Gur-Ari , Ethan Dyer , Behnam Neyshabur

This study aims to understand how statistical biases affect the model's ability to generalize to in-distribution and out-of-distribution data on algorithmic tasks. Prior research indicates that transformers may inadvertently learn to rely…

Machine Learning · Computer Science 2024-09-11 John Mitros

Multilayer transformer networks consist of interleaved self-attention and feedforward sublayers. Could ordering the sublayers in a different pattern lead to better performance? We generate randomly ordered transformers and train them with…

Computation and Language · Computer Science 2020-04-24 Ofir Press , Noah A. Smith , Omer Levy

Despite the widespread success of Transformers on NLP tasks, recent works have found that they struggle to model several formal languages when compared to recurrent models. This raises the question of why Transformers perform well in…

Machine Learning · Computer Science 2023-07-11 Satwik Bhattamishra , Arkil Patel , Varun Kanade , Phil Blunsom

We find limits to the Transformer architecture for language modeling and show it has a universal prediction property in an information-theoretic sense. We further analyze performance in non-asymptotic data regimes to understand the role of…

Machine Learning · Computer Science 2023-07-18 Sourya Basu , Moulik Choraria , Lav R. Varshney

Sequence-to-sequence (seq2seq) models have been successful across many NLP tasks, including ones that require predicting linguistic structure. However, recent work on compositional generalization has shown that seq2seq models achieve very…

Computation and Language · Computer Science 2022-10-25 Yuekun Yao , Alexander Koller

Transformer architectures have facilitated the development of large-scale and general-purpose sequence models for prediction tasks in natural language processing and computer vision, e.g., GPT-3 and Swin Transformer. Although originally…

Machine Learning · Computer Science 2023-06-27 Muning Wen , Runji Lin , Hanjing Wang , Yaodong Yang , Ying Wen , Luo Mai , Jun Wang , Haifeng Zhang , Weinan Zhang

Large language models exhibit surprising emergent generalization properties, yet also struggle on many simple reasoning tasks such as arithmetic and parity. This raises the question of if and when Transformer models can learn the true…

Machine Learning · Computer Science 2023-10-25 Hattie Zhou , Arwen Bradley , Etai Littwin , Noam Razin , Omid Saremi , Josh Susskind , Samy Bengio , Preetum Nakkiran

Large language models (LLMs) based on the Transformer have demonstrated strong performance across diverse tasks. However, current models still exhibit substantial limitations in out-of-distribution (OOD) generalization compared with humans.…

Machine Learning · Computer Science 2026-02-02 Huanyu Liu , Ge Li , Yihong Dong , Sihan Wu , Peixu Wang , Sihao Cheng , Taozhi Chen , Kechi Zhang , Hao Zhu , Tongxuan Liu

Machine translation has seen rapid progress with the advent of Transformer-based models. These models have no explicit linguistic structure built into them, yet they may still implicitly learn structured relationships by attending to…

Recent research suggests that systematic generalization in natural language understanding remains a challenge for state-of-the-art neural models such as Transformers and Graph Neural Networks. To tackle this challenge, we propose Edge…

Computation and Language · Computer Science 2021-12-02 Leon Bergen , Timothy J. O'Donnell , Dzmitry Bahdanau

Compositional generalization is a fundamental trait in humans, allowing us to effortlessly combine known phrases to form novel sentences. Recent works have claimed that standard seq-to-seq models severely lack the ability to compositionally…

Computation and Language · Computer Science 2022-03-16 Arkil Patel , Satwik Bhattamishra , Phil Blunsom , Navin Goyal

Compositional generalization refers to the ability to generalize to novel combinations of previously observed words and syntactic structures. Since it is regarded as a desired property of neural models, recent work has assessed…

Computation and Language · Computer Science 2025-04-07 Ryoma Kumon , Daiki Matsuoka , Hitomi Yanaka

The Transformer model has a tendency to overfit various aspects of the training data, such as the overall sequence length. We study elementary string edit functions using a defined set of error indicators to interpret the behaviour of the…

Machine Learning · Computer Science 2024-10-18 Patrik Zavoral , Dušan Variš , Ondřej Bojar