English
Related papers

Related papers: Truncation Sampling as Language Model Desmoothing

200 papers

Despite their ubiquity in language generation, it remains unknown why truncation sampling heuristics like nucleus sampling are so effective. We provide a theoretical explanation for the effectiveness of the truncation sampling by proving…

Computation and Language · Computer Science 2023-10-04 Matthew Finlayson , John Hewitt , Alexander Koller , Swabha Swayamdipta , Ashish Sabharwal

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

Despite the remarkable advances in language modeling, current mainstream decoding methods still struggle to generate texts that align with human texts across different aspects. In particular, sampling-based methods produce less-repetitive…

Computation and Language · Computer Science 2024-06-06 Haozhe Ji , Pei Ke , Hongning Wang , Minlie Huang

Large Language Models (LLMs) generate text by sampling the next token from a probability distribution over the vocabulary at each decoding step. Popular sampling methods like top-p (nucleus sampling) often struggle to balance quality and…

Computation and Language · Computer Science 2025-11-21 Minh Nhat Nguyen , Andrew Baker , Clement Neo , Allen Roush , Andreas Kirsch , Ravid Shwartz-Ziv

Today's probabilistic language generators fall short when it comes to producing coherent and fluent text despite the fact that the underlying models perform well under standard metrics, e.g., perplexity. This discrepancy has puzzled the…

Computation and Language · Computer Science 2025-06-06 Clara Meister , Tiago Pimentel , Gian Wiher , Ryan Cotterell

Neural text decoding is important for generating high-quality texts using language models. To generate high-quality text, popular decoding algorithms like top-k, top-p (nucleus), and temperature-based sampling truncate or distort the…

Computation and Language · Computer Science 2021-01-18 Sourya Basu , Govardana Sachitanandam Ramachandran , Nitish Shirish Keskar , Lav R. Varshney

Sampling is a common strategy for generating text from probabilistic models, yet standard ancestral sampling often results in text that is incoherent or ungrammatical. To alleviate this issue, various modifications to a model's sampling…

Computation and Language · Computer Science 2024-01-08 Clara Meister , Tiago Pimentel , Luca Malagutti , Ethan G. Wilcox , Ryan Cotterell

In large language model training, input documents are typically concatenated together and then split into sequences of equal length to avoid padding tokens. Despite its efficiency, the concatenation approach compromises data integrity -- it…

Computation and Language · Computer Science 2024-05-03 Hantian Ding , Zijian Wang , Giovanni Paolini , Varun Kumar , Anoop Deoras , Dan Roth , Stefano Soatto

Despite considerable advancements with deep neural language models, the enigma of neural text degeneration persists when these models are tested as text generators. The counter-intuitive empirical observation is that even though the use of…

Computation and Language · Computer Science 2020-02-18 Ari Holtzman , Jan Buys , Li Du , Maxwell Forbes , Yejin Choi

Language models generate text based on successively sampling the next word. A decoding procedure based on nucleus (top-$p$) sampling chooses from the smallest possible set of words whose cumulative probability exceeds the probability $p$.…

Computation and Language · Computer Science 2023-05-05 Shauli Ravfogel , Yoav Goldberg , Jacob Goldberger

The quality of text generated by large language models depends critically on the decoding sampling strategy. While mainstream methods such as Top-$k$, Top-$p$, and Min-$p$ achieve a balance between diversity and accuracy through…

Artificial Intelligence · Computer Science 2026-04-14 Yuanhao Ding , Meimingwei Li , Esteban Garces Arias , Matthias Aßenmacher , Christian Heumann , Chongsheng Zhang

Compute-efficient training of language models has become an important issue. We consider data pruning for data-efficient training of LLMs. In this work, we consider a data pruning method based on information entropy. We propose that the…

Artificial Intelligence · Computer Science 2024-12-13 Minsang Kim , Seungjun Baek

Sampling-based decoding strategies have been widely adopted for Large Language Models (LLMs) in numerous applications, targeting a balance between diversity and quality via temperature tuning and tail truncation. Considering the strong…

Computation and Language · Computer Science 2025-01-09 Yuxuan Zhou , Margret Keuper , Mario Fritz

Top-$k$ decoding is a widely used method for sampling from LLMs: at each token, only the largest $k$ next-token-probabilities are kept, and the next token is sampled after re-normalizing them to sum to unity. Top-$k$ and other sampling…

Artificial Intelligence · Computer Science 2026-02-24 Georgy Noarov , Soham Mallick , Tao Wang , Sunay Joshi , Yan Sun , Yangxinyu Xie , Mengxin Yu , Edgar Dobriban

Given a sequence composed of a limit number of characters, we try to "read" it as a "text". This involves to segment the sequence into "words". The difficulty is to distinguish good segmentation from enormous number of random ones.Aiming at…

Biological Physics · Physics 2009-11-06 Bin Wang

Large language models (LLMs) often struggle to accurately read and comprehend extremely long texts. Current methods for improvement typically rely on splitting long contexts into fixed-length chunks. However, fixed truncation risks…

Computation and Language · Computer Science 2025-06-04 Boheng Sheng , Jiacheng Yao , Meicong Zhang , Guoxiu He

Data deduplication saves storage space by identifying and removing repeats in the data stream. Compared with traditional compression methods, data deduplication schemes are more time efficient and are thus widely used in large scale storage…

Information Theory · Computer Science 2022-05-30 Hao Lou , Farzad Farnoud

Standard decoding strategies for text generation, including top-k, nucleus sampling, and contrastive search, select tokens based on likelihood, restricting selection to high-probability regions. Human language production operates…

Computation and Language · Computer Science 2026-03-20 Esteban Garces Arias , Nurzhan Sapargali , Christian Heumann , Matthias Aßenmacher

In most of neural machine translation distillation or stealing scenarios, the goal is to preserve the performance of the target model (teacher). The highest-scoring hypothesis of the teacher model is commonly used to train a new model…

Computation and Language · Computer Science 2021-04-02 Vilém Zouhar

We find that existing language modeling datasets contain many near-duplicate examples and long repetitive substrings. As a result, over 1% of the unprompted output of language models trained on these datasets is copied verbatim from the…

Computation and Language · Computer Science 2022-03-28 Katherine Lee , Daphne Ippolito , Andrew Nystrom , Chiyuan Zhang , Douglas Eck , Chris Callison-Burch , Nicholas Carlini
‹ Prev 1 2 3 10 Next ›