papers

Publications (14)

cs.LG2024

A Fast, Performant, Secure Distributed Training Framework For Large Language Model

Wei Huang, Yinggui Wang, Anda Cheng +3

The distributed (federated) LLM is an important method for co-training the domain-specific LLM using siloed data. However, maliciously stealing model parameters and data from the s…

cs.LG2021

DPNAS: Neural Architecture Search for Deep Learning with Differential Privacy

Anda Cheng, Jiaxing Wang, Xi Sheryl Zhang +3

Training deep neural networks (DNNs) for meaningful differential privacy (DP) guarantees severely degrades model utility. In this paper, we demonstrate that the architecture of DNN…

cs.LG2026

LLM-AutoDP: Automatic Data Processing via LLM Agents for Model Fine-tuning

Wei Huang, Anda Cheng, Yinggui Wang +2

Large Language Models (LLMs) can be fine-tuned on domain-specific data to enhance their performance in specialized fields. However, such data often contains numerous low-quality sa…

cs.LG2025

DPF-CM: A Data Processing Framework with Privacy-Preserving Vector Databases for Chinese Medical LLMs Training and Deployment

Wei Huang, Anda Cheng, Zhao Zhang +1

Current open-source training pipelines for Chinese medical language models predominantly emphasize optimizing training methodologies to enhance the performance of large language mo…

cs.LG2023

HPN: Personalized Federated Hyperparameter Optimization

Anda Cheng, Zhen Wang, Yaliang Li +1

Numerous research studies in the field of federated learning (FL) have attempted to use personalization to address the heterogeneity among clients, one of FL's most crucial and cha…

cs.CV2022

PKD: General Distillation Framework for Object Detectors via Pearson Correlation Coefficient

Weihan Cao, Yifan Zhang, Jianfei Gao +3

Knowledge distillation(KD) is a widely-used technique to train compact models in object detection. However, there is still a lack of study on how to distill between heterogeneous d…

cs.LG2024

Information Leakage from Embedding in Large Language Models

Zhipeng Wan, Anda Cheng, Yinggui Wang +1

The widespread adoption of large language models (LLMs) has raised concerns regarding data privacy. This study aims to investigate the potential for privacy invasion through input…

cs.CV2019

Location-aware Upsampling for Semantic Segmentation

Xiangyu He, Zitao Mo, Qiang Chen +3

Many successful learning targets such as minimizing dice loss and cross-entropy loss have enabled unprecedented breakthroughs in segmentation tasks. Beyond these semantic metrics,…

cs.CL2026

GradPruner: Gradient-Guided Layer Pruning Enabling Efficient Fine-Tuning and Inference for LLMs

Wei Huang, Anda Cheng, Yinggui Wang

Fine-tuning Large Language Models (LLMs) with downstream data is often considered time-consuming and expensive. Structured pruning methods are primarily employed to improve the inf…

cs.LG2025

Mitigating Catastrophic Forgetting in Large Language Models with Forgetting-aware Pruning

Wei Huang, Anda Cheng, Yinggui Wang

Recent advancements in large language models (LLMs) have shown impressive capabilities in various downstream tasks but typically face Catastrophic Forgetting (CF) during fine-tunin…

cs.LG2022

Differentially Private Federated Learning with Local Regularization and Sparsification

Anda Cheng, Peisong Wang, Xi Sheryl Zhang +1

User-level differential privacy (DP) provides certifiable privacy guarantees to the information that is specific to any user's data in federated learning. Existing methods that ens…

cs.CR2025

CPA-RAG:Covert Poisoning Attacks on Retrieval-Augmented Generation in Large Language Models

Chunyang Li, Junwei Zhang, Anda Cheng +3

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, but its openness introduces vulnerabilities that can be exploited by…

cs.CV2020

SpatialFlow: Bridging All Tasks for Panoptic Segmentation

Qiang Chen, Anda Cheng, Xiangyu He +2

Object location is fundamental to panoptic segmentation as it is related to all things and stuff in the image scene. Knowing the locations of objects in the image provides clues fo…

cs.CR2025

The Shawshank Redemption of Embodied AI: Understanding and Benchmarking Indirect Environmental Jailbreaks

Chunyang Li, Zifeng Kang, Junwei Zhang +4

The adoption of Vision-Language Models (VLMs) in embodied AI agents, while being effective, brings safety concerns such as jailbreaking. Prior work have explored the possibility of…