Author name cluster

Minyi Guo

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

14 papers

2 author rows

ICML Conference 2025 Conference Paper

An Efficient Private GPT Never Autoregressively Decodes

Zhengyi Li 0002
Yue Guan 0003
Kang Yang 0002
Yu Feng 0007
Ning Liu 0007
Yu Yu 0001
Jingwen Leng
Minyi Guo

The wide deployment of the generative pre-trained transformer (GPT) has raised privacy concerns for both clients and servers. While cryptographic primitives can be employed for secure GPT inference to protect the privacy of both parties, they introduce considerable performance overhead. To accelerate secure inference, this study proposes a public decoding and secure verification approach that utilizes public GPT models, motivated by the observation that securely decoding one and multiple tokens takes a similar latency. The client uses the public model to generate a set of tokens, which are then securely verified by the private model for acceptance. The efficiency of our approach depends on the acceptance ratio of tokens proposed by the public model, which we improve from two aspects: (1) a private sampling protocol optimized for cryptographic primitives and (2) model alignment using knowledge distillation. Our approach improves the efficiency of secure decoding while maintaining the same level of privacy and generation quality as standard secure decoding. Experiments demonstrate a $2. 1\times \sim 6. 0\times$ speedup compared to standard decoding across three pairs of public-private models and different network conditions.

Details

NeurIPS Conference 2025 Conference Paper

ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive

Xinhao Luo
Zihan Liu
Yangjie Zhou
Shihan Fang
Ziyu Huang
Yu Feng
Chen Zhang
Shixuan Sun

Large language model (LLM) decoding suffers from high latency due to fragmented execution across operators and heavy reliance on off-chip memory for data exchange and reduction. This execution model limits opportunities for fusion and incurs significant memory traffic and kernel launch overhead. While modern architectures such as NVIDIA Hopper provide distributed shared memory and low-latency intra-cluster interconnects, they expose only low-level data movement instructions, lacking structured abstractions for collective on-chip communication. To bridge this software-hardware gap, we introduce two cluster-level communication primitives, ClusterReduce and ClusterGather, which abstract common communication patterns and enable structured, high-speed data exchange and reduction between thread blocks within a cluster, allowing intermediate results to be on-chip without involving off-chip memory. Building on these abstractions, we design ClusterFusion, an execution framework that schedules communication and computation jointly to expand operator fusion scope by composing decoding stages such as QKV Projection, Attention, and Output Projection into a single fused kernels. Evaluations on H100 GPUs show that ClusterFusion outperforms state-of-the-art inference frameworks by $1. 61\times$ on average in end-to-end latency across different models and configurations.

PDF Details

NeurIPS Conference 2025 Conference Paper

Communication-Efficient Diffusion Denoising Parallelization via Reuse-then-Predict Mechanism

Kunyun Wang
Bohan Li
Kai Yu
Minyi Guo
Jieru Zhao

Diffusion models have emerged as a powerful class of generative models across various modalities, including image, video, and audio synthesis. However, their deployment is often limited by significant inference latency, primarily due to the inherently sequential nature of the denoising process. While existing parallelization strategies attempt to accelerate inference by distributing computation across multiple devices, they typically incur high communication overhead, hindering deployment on commercial hardware. To address this challenge, we propose $\textbf{ParaStep}$, a novel parallelization method based on a reuse-then-predict mechanism that parallelizes diffusion inference by exploiting similarity between adjacent denoising steps. Unlike prior approaches that rely on layer-wise or stage-wise communication, ParaStep employs lightweight, step-wise communication, substantially reducing overhead. ParaStep achieves end-to-end speedups of up to $\textbf{3. 88}$$\times$ on SVD, $\textbf{2. 43}$$\times$ on CogVideoX-2b, and $\textbf{6. 56}$$\times$ on AudioLDM2-large, while maintaining generation quality. These results highlight ParaStep as a scalable and communication-efficient solution for accelerating diffusion inference, particularly in bandwidth-constrained environments.

PDF Details

NeurIPS Conference 2025 Conference Paper

Yggdrasil: Bridging Dynamic Speculation and Static Runtime for Latency-Optimal Tree-Based LLM Decoding

Yue Guan
Changming Yu
Shihan Fang
Weiming Hu
Zaifeng Pan
Zheng Wang
Zihan Liu
Yangjie Zhou

Speculative decoding improves LLM inference by generating and verifying multiple tokens in parallel, but existing systems suffer from suboptimal performance due to a mismatch between dynamic speculation and static runtime assumptions. We present Yggdrasil, a co-designed system that enables latency-optimal speculative decoding through context-aware tree drafting and compiler-friendly execution. Yggdrasil introduces an equal-growth tree structure for static graph compatibility, a latency-aware optimization objective for draft selection, and stage-based scheduling to reduce overhead. Yggdrasil supports unmodified LLMs and achieves up to $3. 98\times$ speedup over state-of-the-art baselines across multiple hardware setups.

PDF Details

NeurIPS Conference 2024 Conference Paper

Nimbus: Secure and Efficient Two-Party Inference for Transformers

Zhengyi Li
Kang Yang
Jin Tan
Wen-jie Lu
Haoqi Wu
Xiao Wang
Yu Yu
Derun Zhao

Transformer models have gained significant attention due to their power in machine learning tasks. Their extensive deployment has raised concerns about the potential leakage of sensitive information during inference. However, when being applied to Transformers, existing approaches based on secure two-party computation (2PC) bring about efficiency limitations in two folds: (1) resource-intensive matrix multiplications in linear layers, and (2) complex non-linear activation functions like $\mathsf{GELU}$ and $\mathsf{Softmax}$. This work presents a new two-party inference framework $\mathsf{Nimbus}$ for Transformer models. Specifically, we propose a new 2PC paradigm to securely compute matrix multiplications based on an outer-product insight, which achieves $2. 9\times \sim 12. 5\times$ performance improvements compared to the state-of-the-art (SOTA) protocol. Furthermore, through a new observation of utilizing the input distribution, we propose an approach of low-degree polynomial approximation for $\mathsf{GELU}$ and $\mathsf{Softmax}$, which improves the performance of the SOTA polynomial approximation by $2. 9\times \sim 4. 0\times$, where the average accuracy loss of our approach is 0. 08\% compared to the non-2PC inference without privacy. Compared with the SOTA two-party inference, $\mathsf{Nimbus}$ improves the end-to-end performance of $BERT_{base}$ inference by $2. 7\times \sim 4. 7\times$ across different network settings.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

SimGen: Simulator-conditioned Driving Scene Generation

Yunsong Zhou
Michael Simon
Zhenghao Peng
Sicheng Mo
Hongzi Zhu
Minyi Guo
Bolei Zhou

Controllable synthetic data generation can substantially lower the annotation cost of training data. Prior works use diffusion models to generate driving images conditioned on the 3D object layout. However, those models are trained on small-scale datasets like nuScenes, which lack appearance and layout diversity. Moreover, overfitting often happens, where the trained models can only generate images based on the layout data from the validation set of the same dataset. In this work, we introduce a simulator-conditioned scene generation framework called SimGen that can learn to generate diverse driving scenes by mixing data from the simulator and the real world. It uses a novel cascade diffusion pipeline to address challenging sim-to-real gaps and multi-condition conflicts. A driving video dataset DIVA is collected to enhance the generative diversity of SimGen, which contains over 147. 5 hours of real-world driving videos from 73 locations worldwide and simulated driving data from the MetaDrive simulator. SimGen achieves superior generation quality and diversity while preserving controllability based on the text prompt and the layout pulled from a simulator. We further demonstrate the improvements brought by SimGen for synthetic data augmentation on the BEV detection and segmentation task and showcase its capability in safety-critical data generation.

PDF Details DOI

IJCAI Conference 2023 Conference Paper

APR: Online Distant Point Cloud Registration through Aggregated Point Cloud Reconstruction

Quan Liu
Yunsong Zhou
Hongzi Zhu
Shan Chang
Minyi Guo

For many driving safety applications, it is of great importance to accurately register LiDAR point clouds generated on distant moving vehicles. However, such point clouds have extremely different point density and sensor perspective on the same object, making registration on such point clouds very hard. In this paper, we propose a novel feature extraction framework, called APR, for online distant point cloud registration. Specifically, APR leverages an autoencoder design, where the autoencoder reconstructs a denser aggregated point cloud with several frames instead of the original single input point cloud. Our design forces the encoder to extract features with rich local geometry information based on one single input point cloud. Such features are then used for online distant point cloud registration. We conduct extensive experiments against state-of-the-art (SOTA) feature extractors on KITTI and nuScenes datasets. Results show that APR outperforms all other extractors by a large margin, increasing average registration recall of SOTA extractors by 7. 1% on LoKITTI and 4. 6% on LoNuScenes. Code is available at https: //github. com/liuQuan98/APR.

PDF Details DOI

AAAI Conference 2022 Conference Paper

Block-Skim: Efficient Question Answering for Transformer

Yue Guan
Zhengyi Li
Zhouhan Lin
Yuhao Zhu
Jingwen Leng
Minyi Guo

Transformer models have achieved promising results on natural language processing (NLP) tasks including extractive question answering (QA). Common Transformer encoders used in NLP tasks process the hidden states of all input tokens in the context paragraph throughout all layers. However, different from other tasks such as sequence classification, answering the raised question does not necessarily need all the tokens in the context paragraph. Following this motivation, we propose Block-Skim, which learns to skim unnecessary context in higher hidden layers to improve and accelerate the Transformer performance. The key idea of Block-Skim is to identify the context that must be further processed and those that could be safely discarded early on during inference. Critically, we find that such information could be sufficiently derived from the self-attention weights inside the Transformer model. We further prune the hidden states corresponding to the unnecessary positions early in lower layers, achieving significant inference-time speedup. To our surprise, we observe that models pruned in this way outperform their full-size counterparts. Block-Skim improves QA models’ accuracy on different datasets and achieves 3× speedup on BERTbase model.

PDF Details

NeurIPS Conference 2022 Conference Paper

MoGDE: Boosting Mobile Monocular 3D Object Detection with Ground Depth Estimation

Yunsong Zhou
Quan Liu
Hongzi Zhu
Yunzhe Li
Shan Chang
Minyi Guo

Monocular 3D object detection (Mono3D) in mobile settings (e. g. , on a vehicle, a drone, or a robot) is an important yet challenging task. Due to the near-far disparity phenomenon of monocular vision and the ever-changing camera pose, it is hard to acquire high detection accuracy, especially for far objects. Inspired by the insight that the depth of an object can be well determined according to the depth of the ground where it stands, in this paper, we propose a novel Mono3D framework, called MoGDE, which constantly estimates the corresponding ground depth of an image and then utilizes the estimated ground depth information to guide Mono3D. To this end, we utilize a pose detection network to estimate the pose of the camera and then construct a feature map portraying pixel-level ground depth according to the 3D-to-2D perspective geometry. Moreover, to improve Mono3D with the estimated ground depth, we design an RGB-D feature fusion network based on the transformer structure, where the long-range self-attention mechanism is utilized to effectively identify ground-contacting points and pin the corresponding ground depth to the image feature map. We conduct extensive experiments on the real-world KITTI dataset. The results demonstrate that MoGDE can effectively improve the Mono3D accuracy and robustness for both near and far objects. MoGDE yields the best performance compared with the state-of-the-art methods by a large margin and is ranked number one on the KITTI 3D benchmark.

PDF Details

ICLR Conference 2022 Conference Paper

SQuant: On-the-Fly Data-Free Quantization via Diagonal Hessian Approximation

Cong Guo 0003
Yuxian Qiu
Jingwen Leng
Xiaotian Gao
Chen Zhang 0001
Yunxin Liu 0001
Fan Yang 0024
Yuhao Zhu 0001

Quantization of deep neural networks (DNN) has been proven effective for compressing and accelerating DNN models. Data-free quantization (DFQ) is a promising approach without the original datasets under privacy-sensitive and confidential scenarios. However, current DFQ solutions degrade accuracy, need synthetic data to calibrate networks, and are time-consuming and costly. This paper proposes an on-the-fly DFQ framework with sub-second quantization time, called SQuant, which can quantize networks on inference-only devices with low computation and memory requirements. With the theoretical analysis of the second-order information of DNN task loss, we decompose and approximate the Hessian-based optimization objective into three diagonal sub-items, which have different areas corresponding to three dimensions of weight tensor: element-wise, kernel-wise, and output channel-wise. Then, we progressively compose sub-items and propose a novel data-free optimization objective in the discrete domain, minimizing Constrained Absolute Sum of Error (or CASE in short), which surprisingly does not need any dataset and is even not aware of network architecture. We also design an efficient algorithm without back-propagation to further reduce the computation complexity of the objective solver. Finally, without fine-tuning and synthetic datasets, SQuant accelerates the data-free quantization process to a sub-second level with >30% accuracy improvement over the existing data-free post-training quantization works, with the evaluated models under 4-bit quantization. We have open-sourced the SQuant framework at https://github.com/clevercool/SQuant.

Details

AAAI Conference 2018 Conference Paper

Deep Representation-Decoupling Neural Networks for Monaural Music Mixture Separation

Zhuo Li
Hongwei Wang
Miao Zhao
Wenjie Li
Minyi Guo

Monaural source separation (MSS) aims to extract and reconstruct different sources from a single-channel mixture, which could facilitate a variety of applications such as chord recognition, pitch estimation and automatic transcription. In this paper, we study the problem of separating vocals and instruments from monaural music mixture. Existing works for monaural source separation either utilize linear and shallow models (e. g. , non-negative matrix factorization), or do not explicitly address the coupling and tangling of multiple sources in original input signals, hence they do not perform satisfactorily in real-world scenarios. To overcome the above limitations, we propose a novel end-to-end framework for monaural music mixture separation called Deep Representation- Decoupling Neural Networks (DRDNN). DRDNN takes advantages of both traditional signal processing methods and popular deep learning models. For each input of music mixture, DRDNN converts it to a two-dimensional timefrequency spectrogram using short-time Fourier transform (STFT), followed by stacked convolutional neural networks (CNN) layers and long-short term memory (LSTM) layers to extract more condensed features. Afterwards, DRDNN utilizes a decoupling component, which consists of a group of multi-layer perceptrons (MLP), to decouple the features further into different separated sources. The design of decoupling component in DRDNN produces puriﬁed single-source signals for subsequent full-size restoration, and can signiﬁcantly improve the performance of ﬁnal separation. Through extensive experiments on real-world dataset, we prove that DRDNN outperforms state-of-the-art baselines in the task of monaural music mixture separation and reconstruction.

PDF Details

AAAI Conference 2018 Conference Paper

GraphGAN: Graph Representation Learning With Generative Adversarial Nets

Hongwei Wang
Jia Wang
Jialin Wang
Miao Zhao
Weinan Zhang
Fuzheng Zhang
Xing Xie
Minyi Guo

The goal of graph representation learning is to embed each vertex in a graph into a low-dimensional vector space. Existing graph representation learning methods can be classiﬁed into two categories: generative models that learn the underlying connectivity distribution in the graph, and discriminative models that predict the probability of edge existence between a pair of vertices. In this paper, we propose Graph- GAN, an innovative graph representation learning framework unifying above two classes of methods, in which the generative model and discriminative model play a game-theoretical minimax game. Speciﬁcally, for a given vertex, the generative model tries to ﬁt its underlying true connectivity distribution over all other vertices and produces “fake” samples to fool the discriminative model, while the discriminative model tries to detect whether the sampled vertex is from ground truth or generated by the generative model. With the competition between these two models, both of them can alternately and iteratively boost their performance. Moreover, when considering the implementation of generative model, we propose a novel graph softmax to overcome the limitations of traditional softmax function, which can be proven satisfying desirable properties of normalization, graph structure awareness, and computational efﬁciency. Through extensive experiments on real-world datasets, we demonstrate that Graph- GAN achieves substantial gains in a variety of applications, including link prediction, node classiﬁcation, and recommendation, over state-of-the-art baselines.

PDF Details

TAAS Journal 2014 Journal Article

Modeling and Defending against Adaptive BitTorrent Worms in Peer-to-Peer Networks

Jiaqing Luo
Bin Xiao
Qingjun Xiao
Jiannong Cao
Minyi Guo

BitTorrent (BT) is one of the most common Peer-to-Peer (P2P) file sharing protocols. Rather than downloading a file from a single source, the protocol allows users to join a swarm of peers to download and upload from each other simultaneously. Worms exploiting information from BT servers or trackers can cause serious damage to participating peers, which unfortunately has been neglected previously. In this article, we first present a new worm, called Adaptive BitTorrent worm (A-BT worm), which finds new victims and propagates sending forged requests to trackers. To reduce its abnormal behavior, the worm estimates the ratio of infected peers and adaptively adjusts its propagation speed. We then build a hybrid model to precisely characterize the propagation behavior of the worm. We also propose a statistical method to automatically detect the worm from the tracker by estimating the variance of the time intervals of requests. To slow down the worm propagation, we design a safe strategy in which the tracker returns secured peers when receives a request. Finally, we evaluate the accuracy of the hybrid model, and the effectiveness of our detection method and containment strategy through simulations.

Details DOI

AAAI Conference 2012 Conference Paper

Emoticon Smoothed Language Models for Twitter Sentiment Analysis

Kun-Lin Liu
Wu-Jun Li
Minyi Guo

Twitter sentiment analysis (TSA) has become a hot research topic in recent years. The goal of this task is to discover the attitude or opinion of the tweets, which is typically formulated as a machine learning based text classification problem. Some methods use manually labeled data to train fully supervised models, while others use some noisy labels, such as emoticons and hashtags, for model training. In general, we can only get a limited number of training data for the fully supervised models because it is very labor-intensive and time-consuming to manually label the tweets. As for the models with noisy labels, it is hard for them to achieve satisfactory performance due to the noise in the labels although it is easy to get a large amount of data for training. Hence, the best strategy is to utilize both manually labeled data and noisy labeled data for training. However, how to seamlessly integrate these two different kinds of data into the same learning framework is still a challenge. In this paper, we present a novel model, called emoticon smoothed language model (ESLAM), to handle this challenge. The basic idea is to train a language model based on the manually labeled data, and then use the noisy emoticon data for smoothing. Experiments on real data sets demonstrate that ESLAM can effectively integrate both kinds of data to outperform those methods using only one of them.

PDF Details