English
Related papers

Related papers: Temperature-Centric Investigation of Speculative D…

200 papers

Transformer-based autoregressive sampling has been the major bottleneck for slowing down large language model inferences. One effective way to accelerate inference is \emph{Speculative Decoding}, which employs a small model to sample a…

Machine Learning · Computer Science 2024-11-05 Ming Yin , Minshuo Chen , Kaixuan Huang , Mengdi Wang

Knowledge Distillation (KD) is one proposed solution to large model sizes and slow inference speed in semantic segmentation. In our research we identify 25 proposed distillation loss terms from 14 publications in the last 4 years.…

Computer Vision and Pattern Recognition · Computer Science 2023-09-08 Onno Niemann , Christopher Vox , Thorben Werner

As a technique to bridge logit matching and probability distribution matching, temperature scaling plays a pivotal role in knowledge distillation (KD). Conventionally, temperature scaling is applied to both teacher's logits and student's…

Machine Learning · Computer Science 2024-03-11 Kaixiang Zheng , En-Hui Yang

Decoding from the output distributions of large language models to produce high-quality text is a complex challenge in language modeling. Various approaches, such as beam search, sampling with temperature, $k-$sampling, nucleus…

Computation and Language · Computer Science 2024-10-22 Esteban Garces Arias , Julian Rodemann , Meimingwei Li , Christian Heumann , Matthias Aßenmacher

Knowledge distillation (KD) is widely used for training a compact model with the supervision of another large model, which could effectively improve the performance. Previous methods mainly focus on two aspects: 1) training the student to…

Computer Vision and Pattern Recognition · Computer Science 2020-07-27 Tiancheng Wen , Shenqi Lai , Xueming Qian

Spatiotemporal forecasting often relies on computationally intensive models to capture complex dynamics. Knowledge distillation (KD) has emerged as a key technique for creating lightweight student models, with recent advances like…

Machine Learning · Computer Science 2025-12-02 Wenshuo Wang , Yaomin Shen , Yingjie Tan , Yihao Chen

Speculative decoding has emerged as a pivotal technique to accelerate LLM inference by employing a lightweight draft model to generate candidate tokens that are subsequently verified by the target model in parallel. However, while this…

Computation and Language · Computer Science 2026-02-26 Yuetao Chen , Xuliang Wang , Xinzhou Zheng , Ming Li , Peng Wang , Hong Xu

Speculative decoding accelerates LLM inference by using a fast draft model to generate tokens and a more accurate target model to verify them. Its performance depends on the $\textit{acceptance length}$, or number of draft tokens accepted…

Computation and Language · Computer Science 2026-05-20 Avinash Kumar , Sujay Sanghavi , Poulami Das

LLMs are increasingly explored for bundle generation, thanks to their reasoning capabilities and knowledge. However, deploying large-scale LLMs introduces significant efficiency challenges, primarily high computational costs during…

Computation and Language · Computer Science 2025-04-25 Kaidong Feng , Zhu Sun , Jie Yang , Hui Fang , Xinghua Qu , Wenyuan Liu

Deployed large language models (LLMs) often rely on speculative decoding, a technique that generates and verifies multiple candidate tokens in parallel, to improve throughput and latency. In this work, we reveal a new side-channel whereby…

Computation and Language · Computer Science 2026-02-12 Jiankun Wei , Abdulrahman Abdulrazzag , Tianchen Zhang , Adel Muursepp , Gururaj Saileshwar

Large language models (LLMs) can improve reasoning at inference time through test-time scaling (TTS), where multiple reasoning traces are generated and the best one is selected. Prior work shows that increasing the number of samples K…

Artificial Intelligence · Computer Science 2025-10-06 Yuheng Wu , Azalia Mirhoseini , Thierry Tambe

We propose a novel speculative decoding method tailored for multi-sample reasoning scenarios, such as self-consistency and Best-of-N sampling. Our method exploits the intrinsic consensus of parallel generation paths to synthesize…

Computation and Language · Computer Science 2025-03-10 Yiwei Li , Jiayi Shi , Shaoxiong Feng , Peiwen Yuan , Xinglin Wang , Yueqi Zhang , Ji Zhang , Chuyi Tan , Boyuan Pan , Yao Hu , Kan Li

Knowledge Distillation (KD) aims to transfer knowledge from a large teacher model to a smaller student model. While contrastive learning has shown promise in self-supervised learning by creating discriminative representations, its…

Computer Vision and Pattern Recognition · Computer Science 2025-05-14 Nikolaos Giakoumoglou , Tania Stathaki

Knowledge distillation (KD) is a promising technique for model compression in neural machine translation. However, where the knowledge hides in KD is still not clear, which may hinder the development of KD. In this work, we first unravel…

Computation and Language · Computer Science 2024-07-18 Songming Zhang , Yunlong Liang , Shuaibo Wang , Wenjuan Han , Jian Liu , Jinan Xu , Yufeng Chen

This paper introduces Multimodal Speculative Decoding (MSD) to accelerate Multimodal Large Language Models (MLLMs) inference. Speculative decoding has been shown to accelerate Large Language Models (LLMs) without sacrificing accuracy.…

Computer Vision and Pattern Recognition · Computer Science 2025-05-21 Luxi Lin , Zhihang Lin , Zhanpeng Zeng , Rongrong Ji

Generative models of complex systems often require post-hoc parameter adjustments to produce useful outputs. For example, energy-based models for protein design are sampled at an artificially low ''temperature'' to generate novel,…

Quantitative Methods · Quantitative Biology 2025-12-11 Peter W Fields , Vudtiwat Ngampruetikorn , David J Schwab , Stephanie E Palmer

Reasoning models excel by generating long chain-of-thoughts, but decoding the resulting thousands of tokens is slow. Token-level speculative decoding (SD) helps, but its benefit is capped, because the chance that an entire $\gamma$-token…

Machine Learning · Computer Science 2025-06-25 Yichao Fu , Rui Ge , Zelei Shao , Zhijie Deng , Hao Zhang

Several recent studies have elucidated why knowledge distillation (KD) improves model performance. However, few have researched the other advantages of KD in addition to its improving model performance. In this study, we have attempted to…

Machine Learning · Computer Science 2023-05-26 Hyeongrok Han , Siwon Kim , Hyun-Soo Choi , Sungroh Yoon

Speculative decoding is a powerful technique that attempts to circumvent the autoregressive constraint of modern Large Language Models (LLMs). The aim of speculative decoding techniques is to improve the average inference time of a large,…

Computation and Language · Computer Science 2024-10-25 Sudhanshu Agrawal , Wonseok Jeon , Mingu Lee

Autoregressive decoding is bottlenecked by its sequential nature. Speculative decoding has become a standard way to accelerate inference by using a fast draft model to predict upcoming tokens from a slower target model, and then verifying…

Machine Learning · Computer Science 2026-05-06 Tanishq Kumar , Tri Dao , Avner May