English
Related papers

Related papers: PaSS: Parallel Speculative Sampling

200 papers

The quadratic complexity of attention imposes severe memory and computational bottlenecks on Large Language Model (LLM) inference. This challenge is particularly acute for emerging agentic applications that require processing multi-million…

Machine Learning · Computer Science 2026-05-19 Ceyu Xu , Jiangnan Yu , Yongji Wu , Yuan Xie

As the basis of generative AI, an autoregressive model requires the generation of a new token depending on all the previously generated tokens, which brings high quality but also restricts the model to generate tokens one by one, forming a…

Computation and Language · Computer Science 2025-07-02 Zixian Huang , Chenxu Niu , Yu Gu , Gengyang Xiao , Xinwei Huang , Gong Cheng

As large language models gain widespread adoption, running them efficiently becomes crucial. Recent works on LLM inference use speculative decoding to achieve extreme speedups. However, most of these works implicitly design their algorithms…

Computation and Language · Computer Science 2024-12-03 Ruslan Svirschevski , Avner May , Zhuoming Chen , Beidi Chen , Zhihao Jia , Max Ryabinin

The autoregressive nature of large language models (LLMs) fundamentally limits inference speed, as each forward pass generates only a single token and is often bottlenecked by memory bandwidth. Speculative decoding has emerged as a…

Machine Learning · Computer Science 2025-12-02 Zihao An , Huajun Bai , Ziqiong Liu , Dong Li , Emad Barsoum

Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast…

Computation and Language · Computer Science 2026-05-29 Jian Chen , Yesheng Liang , Zhijian Liu

While sequence-to-sequence (seq2seq) models achieve state-of-the-art performance in many natural language processing tasks, they can be too slow for real-time applications. One performance bottleneck is predicting the most likely next token…

Computation and Language · Computer Science 2019-07-26 Chunyang Xiao , Christoph Teichmann , Konstantine Arkoudas

Modern autoregressive speech synthesis models leveraging language models have demonstrated remarkable performance. However, the sequential nature of next token prediction in these models leads to significant latency, hindering their…

Sound · Computer Science 2025-06-04 Zijian Lin , Yang Zhang , Yougen Yuan , Yuming Yan , Jinjiang Liu , Zhiyong Wu , Pengfei Hu , Qun Yu

Speculative decoding (SD) is a powerful technique for accelerating the inference process of large language models (LLMs) without sacrificing accuracy. Typically, SD employs a small draft model to generate a fixed number of draft tokens,…

Computation and Language · Computer Science 2026-02-03 Situo Zhang , Yifan Zhang , Zichen Zhu , Hankun Wang , Da Ma , Danyang Zhang , Lu Chen , Kai Yu

Speculative decoding accelerates memory-bound LLM inference without quality degradation by using a fast drafter to propose multiple candidate tokens and the target model to verify them in parallel. However, conventional sequential…

Computation and Language · Computer Science 2026-05-20 Yaojie Zhang , Jianuo Huang , Junlong Ke , Yuhang Han , Yongji Long , Tianchen Zhao , Biqing Qi , Linfeng Zhang

Speculative decoding has emerged as a promising technique to accelerate the inference of Large Language Models (LLMs) by employing a small language model to draft a hypothesis sequence, which is then validated by the LLM. The effectiveness…

Computation and Language · Computer Science 2024-07-24 Zhuocheng Gong , Jiahao Liu , Ziyue Wang , Pengfei Wu , Jingang Wang , Xunliang Cai , Dongyan Zhao , Rui Yan

End-to-end automatic speech recognition (ASR) systems based on transformer architectures, such as Whisper, offer high transcription accuracy and robustness. However, their autoregressive decoding is computationally expensive, hence limiting…

Computation and Language · Computer Science 2025-07-30 Tuan Vu Ho , Hiroaki Kokubo , Masaaki Yamamoto , Yohei Kawaguchi

Speculative decoding (SD) accelerates Large Language Model (LLM) generation by using an efficient draft model to propose the next few tokens, which are verified by the LLM in a single forward call, reducing latency while preserving its…

Computation and Language · Computer Science 2025-05-30 Milan Gritta , Huiyin Xue , Gerasimos Lampouras

Recent advances in visual generation have made significant strides in producing content of exceptional quality. However, most methods suffer from a fundamental problem - a bottleneck of inference computational efficiency. Most of these…

Computer Vision and Pattern Recognition · Computer Science 2025-02-04 Sahil Goyal , Debapriya Tula , Gagan Jain , Pradeep Shenoy , Prateek Jain , Sujoy Paul

Inference with Multimodal Large Language Models (MLLMs) is slow due to their large-language-model backbone which suffers from memory bandwidth bottleneck and generates tokens auto-regressively. In this paper, we explore the application of…

Computation and Language · Computer Science 2024-04-16 Mukul Gagrani , Raghavv Goel , Wonseok Jeon , Junyoung Park , Mingu Lee , Christopher Lott

Speculative Decoding has gained popularity as an effective technique for accelerating the auto-regressive inference process of Large Language Models. However, Speculative Decoding entirely relies on the availability of efficient draft…

Computation and Language · Computer Science 2025-06-06 Ofir Zafrir , Igor Margulis , Dorin Shteyman , Shira Guskin , Guy Boudoukh

Model ensembling is a technique to combine the predicted distributions of two or more models, often leading to improved robustness and performance. For ensembling in text generation, the next token's probability distribution is derived from…

Computation and Language · Computer Science 2025-03-03 Rachel Wicks , Kartik Ravisankar , Xinchen Yang , Philipp Koehn , Matt Post

Large language models achieve impressive performance across diverse tasks but exhibit high inference latency due to their large parameter sizes. While quantization reduces model size, it often leads to performance degradation compared to…

Hardware Architecture · Computer Science 2025-10-22 Yushu Zhao , Yubin Qin , Yang Wang , Xiaolong Yang , Huiming Han , Shaojun Wei , Yang Hu , Shouyi Yin

The autoregressive nature of conventional large language models (LLMs) inherently limits inference speed, as tokens are generated sequentially. While speculative and parallel decoding techniques attempt to mitigate this, they face…

Artificial Intelligence · Computer Science 2024-10-22 Aishwarya P S , Pranav Ajit Nair , Yashas Samaga , Toby Boyd , Sanjiv Kumar , Prateek Jain , Praneeth Netrapalli

Auto-regressive models are widely used in sequence generation problems. The output sequence is typically generated in a predetermined order, one discrete unit (pixel or word or character) at a time. The models are trained by teacher-forcing…

Computation and Language · Computer Science 2019-10-23 Daniel Duckworth , Arvind Neelakantan , Ben Goodrich , Lukasz Kaiser , Samy Bengio

In this paper we propose a novel method of augmenting parallel text corpora which promises good quality and is also capable of producing many fold larger corpora than the seed corpus we start with. We do not need any additional monolingual…

Computation and Language · Computer Science 2024-10-07 Vibhuti Kumari , Narayana Murthy Kavi
‹ Prev 1 4 5 6 7 8 10 Next ›