English
Related papers

Related papers: Text Generation Beyond Discrete Token Sampling

200 papers

Auto-regressive sequence generative models trained by Maximum Likelihood Estimation suffer the exposure bias problem in practical finite sample scenarios. The crux is that the number of training samples for Maximum Likelihood Estimation is…

Machine Learning · Statistics 2020-07-14 Yuxuan Song , Ning Miao , Hao Zhou , Lantao Yu , Mingxuan Wang , Lei Li

Controlled text generation allows for enforcing user-defined constraints on large language model outputs, an increasingly important field as LLMs become more prevalent in everyday life. One common approach uses energy-based decoding, which…

Computation and Language · Computer Science 2025-02-07 Patrick Pynadath , Ruqi Zhang

Autoregressive language models are constrained by their inherently sequential nature, generating one token at a time. This paradigm limits inference speed and parallelism, especially during later stages of generation when the direction and…

Computation and Language · Computer Science 2025-07-17 Mohammad Samragh , Arnav Kundu , David Harrison , Kumari Nishu , Devang Naik , Minsik Cho , Mehrdad Farajtabar

Standard autoregressive language models generate text by repeatedly selecting a discrete next token, coupling prediction with irreversible commitment at every step. We show that token selection is not the only viable autoregressive…

Computation and Language · Computer Science 2026-04-07 Oshri Naparstek

Most sequence-to-sequence (seq2seq) models are autoregressive; they generate each token by conditioning on previously generated tokens. In contrast, non-autoregressive seq2seq models generate all tokens in one pass, which leads to increased…

Computation and Language · Computer Science 2019-10-10 Xuezhe Ma , Chunting Zhou , Xian Li , Graham Neubig , Eduard Hovy

A recent study showed that large language models (LLMs) can reconstruct surprisingly long texts - up to thousands of tokens - via autoregressive generation from just one trained input embedding. In this work, we explore whether…

Computation and Language · Computer Science 2025-11-04 Gleb Mezentsev , Ivan Oseledets

The input and output of most text generation tasks can be transformed to two sequences of tokens and they can be modeled using sequence-to-sequence learning modeling tools such as Transformers. These models are usually trained by maximizing…

Computation and Language · Computer Science 2022-12-09 Xingxing Zhang , Yiran Liu , Xun Wang , Pengcheng He , Yang Yu , Si-Qing Chen , Wayne Xiong , Furu Wei

This paper argues that generating output tokens is more effective than using pooled representations for prediction tasks because token-level generation retains more mutual information. Since LLMs are trained on massive text corpora using…

Autoregressive language models accumulate errors due to their fixed, irrevocable left-to-right token generation. To address this, we propose a new sampling method called Resample-Previous-Tokens (RPT). RPT mitigates error accumulation by…

Machine Learning · Computer Science 2025-06-09 Itai Gat , Neta Shaul , Uriel Singer , Yaron Lipman

The autoregressive decoding for text generation in large language models (LLMs), while widely used, is inherently suboptimal due to the lack of a built-in mechanism to perform refinement and/or correction of the generated content. In this…

Computation and Language · Computer Science 2025-06-03 Zeyu Tang , Zhenhao Chen , Xiangchen Song , Loka Li , Yunlong Deng , Yifan Shen , Guangyi Chen , Peter Spirtes , Kun Zhang

Retrieval-augmented generation (RAG) has emerged as a popular approach to steering the output of a large language model (LLM) by incorporating retrieved contexts as inputs. However, existing work observed the generator bias, such that…

Computation and Language · Computer Science 2024-12-17 Youngwon Lee , Seung-won Hwang , Daniel Campos , Filip Graliński , Zhewei Yao , Yuxiong He

Traditional language models operate autoregressively, i.e., they predict one token at a time. Rapid explosion in model sizes has resulted in high inference times. In this work, we propose DynaMo, a suite of multi-token prediction language…

Computation and Language · Computer Science 2024-05-03 Shikhar Tuli , Chi-Heng Lin , Yen-Chang Hsu , Niraj K. Jha , Yilin Shen , Hongxia Jin

LLMs have become the mainstream approaches to code generation. Existing LLMs mainly employ autoregressive generation, i.e. generating code token-by-token from left to right. However, the underlying autoregressive generation has two…

Software Engineering · Computer Science 2025-11-04 Chengze Li , Yitong Zhang , Jia Li , Liyi Cai , Ge Li

Developing a generative model of realistic order flow in financial markets is a challenging open problem, with numerous applications for market participants. Addressing this, we propose the first end-to-end autoregressive generative model…

Trading and Market Microstructure · Quantitative Finance 2023-09-06 Peer Nagy , Sascha Frey , Silvia Sapora , Kang Li , Anisoara Calinescu , Stefan Zohren , Jakob Foerster

Standard sequential generation methods assume a pre-specified generation order, such as text generation methods which generate words from left to right. In this work, we propose a framework for training models of text generation that…

Computation and Language · Computer Science 2019-10-25 Sean Welleck , Kianté Brantley , Hal Daumé , Kyunghyun Cho

Diffusion models have emerged as a dominant paradigm for generative modeling across a wide range of domains, including prompt-conditional generation. The vast majority of samplers, however, rely on forward discretization of the reverse…

Computer Vision and Pattern Recognition · Computer Science 2025-12-01 Zhenghan Fang , Jian Zheng , Qiaozi Gao , Xiaofeng Gao , Jeremias Sulam

Diffusion language models (DLMs) promise parallel, order-agnostic generation, but on standard benchmarks they have historically lagged behind autoregressive models in sample quality and diversity. Recent continuous flow and diffusion…

Computation and Language · Computer Science 2026-05-11 Georgios Batzolis , Mark Girolami , Luca Ambrogioni

Despite their growing capabilities, language models still frequently reproduce content from their training data, generate repetitive text, and favor common grammatical patterns and vocabulary. A possible cause is the decoding strategy: the…

Computation and Language · Computer Science 2026-01-15 Giorgio Franceschelli , Mirco Musolesi

Recent endeavors in Multimodal Large Language Models (MLLMs) aim to unify visual comprehension and generation by combining LLM and diffusion models, the state-of-the-art in each task, respectively. Existing approaches rely on spatial visual…

Computer Vision and Pattern Recognition · Computer Science 2025-04-22 Kaihang Pan , Wang Lin , Zhongqi Yue , Tenglong Ao , Liyu Jia , Wei Zhao , Juncheng Li , Siliang Tang , Hanwang Zhang

Autoregressive (AR) Large Language Models (LLMs) have demonstrated significant success across numerous tasks. However, the AR modeling paradigm presents certain limitations; for instance, contemporary autoregressive LLMs are trained to…

Machine Learning · Computer Science 2025-02-10 Justin Deschenaux , Caglar Gulcehre
‹ Prev 1 2 3 10 Next ›