English
Related papers

Related papers: Token-Level Fitting Issues of Seq2seq Models

200 papers

Large language models are pre-trained on ever-growing token budgets under the assumption that better pre-training performance translates to improved downstream models. In this work, we challenge this assumption and show that extended…

Computation and Language · Computer Science 2025-03-31 Jacob Mitchell Springer , Sachin Goyal , Kaiyue Wen , Tanishq Kumar , Xiang Yue , Sadhika Malladi , Graham Neubig , Aditi Raghunathan

In this paper, we present a substantial step in better understanding the SOTA sequence-to-sequence (Seq2Seq) pretraining for neural machine translation~(NMT). We focus on studying the impact of the jointly pretrained decoder, which is the…

Computation and Language · Computer Science 2022-03-17 Wenxuan Wang , Wenxiang Jiao , Yongchang Hao , Xing Wang , Shuming Shi , Zhaopeng Tu , Michael Lyu

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

In language processing, training data with extremely large variance may lead to difficulty in the language model's convergence. It is difficult for the network parameters to adapt sentences with largely varied semantics or grammatical…

Computation and Language · Computer Science 2022-05-26 Yunhao Yang , Zhaokun Xue

Neural language models are a critical component of state-of-the-art systems for machine translation, summarization, audio transcription, and other tasks. These language models are almost universally autoregressive in nature, generating…

Machine Learning · Computer Science 2018-08-27 Nicolas Ford , Daniel Duckworth , Mohammad Norouzi , George E. Dahl

Recurrent neural networks are nowadays successfully used in an abundance of applications, going from text, speech and image processing to recommender systems. Backpropagation through time is the algorithm that is commonly used to train…

Machine Learning · Computer Science 2018-01-10 Cedric De Boom , Thomas Demeester , Bart Dhoedt

Neural sequence-to-sequence models are currently the dominant approach in several natural language processing tasks, but require large parallel corpora. We present a sequence-to-sequence-to-sequence autoencoder (SEQ^3), consisting of two…

Computation and Language · Computer Science 2019-06-11 Christos Baziotis , Ion Androutsopoulos , Ioannis Konstas , Alexandros Potamianos

Sequence-to-Sequence models were introduced to tackle many real-life problems like machine translation, summarization, image captioning, etc. The standard optimization algorithms are mainly based on example-to-example matching like maximum…

Computation and Language · Computer Science 2018-09-05 Wenhu Chen , Guanlin Li , Shujie Liu , Zhirui Zhang , Mu Li , Ming Zhou

Large language models have demonstrated predictable scaling behaviors with respect to model parameters and training data. This study investigates whether a similar scaling relationship exist for vision-language models with respect to the…

Artificial Intelligence · Computer Science 2025-12-30 Tenghui Li , Guoxu Zhou , Xuyang Zhao , Qibin Zhao

Sentence simplification is the task of rewriting texts so they are easier to understand. Recent research has applied sequence-to-sequence (Seq2Seq) models to this task, focusing largely on training-time improvements via reinforcement…

Computation and Language · Computer Science 2019-04-08 Reno Kriz , João Sedoc , Marianna Apidianaki , Carolina Zheng , Gaurav Kumar , Eleni Miltsakaki , Chris Callison-Burch

Pretraining large language models effectively requires strategic data selection, blending and ordering. However, key details about data mixtures especially their scalability to longer token horizons and larger model sizes remain…

Computation and Language · Computer Science 2024-12-23 Steven Feng , Shrimai Prabhumoye , Kezhi Kong , Dan Su , Mostofa Patwary , Mohammad Shoeybi , Bryan Catanzaro

Sequence-to-sequence (seq2seq) models are prevalent in semantic parsing, but have been found to struggle at out-of-distribution compositional generalization. While specialized model architectures and pre-training of seq2seq models have been…

Computation and Language · Computer Science 2021-04-16 Jonathan Herzig , Peter Shaw , Ming-Wei Chang , Kelvin Guu , Panupong Pasupat , Yuan Zhang

This paper introduces the counter-intuitive generalization results of overfitting pre-trained large language models (LLMs) on very small datasets. In the setting of open-ended text generation, it is well-documented that LLMs tend to…

Computation and Language · Computer Science 2025-02-27 Fredrik Carlsson , Fangyu Liu , Daniel Ward , Murathan Kurfali , Joakim Nivre

Crafting adversarial examples has become an important technique to evaluate the robustness of deep neural networks (DNNs). However, most existing works focus on attacking the image classification problem since its input space is continuous…

Machine Learning · Computer Science 2020-04-22 Minhao Cheng , Jinfeng Yi , Pin-Yu Chen , Huan Zhang , Cho-Jui Hsieh

Backdoor learning has become an emerging research area towards building a trustworthy machine learning system. While a lot of works have studied the hidden danger of backdoor attacks in image or text classification, there is a limited…

Computation and Language · Computer Science 2023-05-05 Lichang Chen , Minhao Cheng , Heng Huang

Large language models make remarkable progress in reasoning capabilities. Existing works focus mainly on deductive reasoning tasks (e.g., code and math), while another type of reasoning mode that better aligns with human learning, inductive…

Computation and Language · Computer Science 2025-03-18 Kedi Chen , Zhikai Lei , Fan Zhang , Yinqi Zhang , Qin Chen , Jie Zhou , Liang He , Qipeng Guo , Kai Chen , Wei Zhang

Modern large language models use a fixed tokenizer to effectively compress text drawn from a source domain. However, applying the same tokenizer to a new target domain often leads to inferior compression, more costly inference, and reduced…

Computation and Language · Computer Science 2024-11-07 Zhili Feng , Tanya Marwah , Nicolo Fusi , David Alvarez-Melis , Lester Mackey

Attention mechanism is a fundamental component of the transformer model and plays a significant role in its success. However, the theoretical understanding of how attention learns to select tokens is still an emerging area of research. In…

Machine Learning · Computer Science 2025-05-20 Keitaro Sakamoto , Issei Sato

Attention-based models trained on protein sequences have demonstrated incredible success at classification and generation tasks relevant for artificial intelligence-driven protein design. However, we lack a sufficient understanding of how…

Machine Learning · Computer Science 2022-06-29 Erik Nijkamp , Jeffrey Ruffolo , Eli N. Weinstein , Nikhil Naik , Ali Madani

Pre-trained encoder-only and sequence-to-sequence (seq2seq) models each have advantages, however training both model types from scratch is computationally expensive. We explore recipes to improve pre-training efficiency by initializing one…

Computation and Language · Computer Science 2023-06-16 Saleh Soltan , Andy Rosenbaum , Tobias Falke , Qin Lu , Anna Rumshisky , Wael Hamza