Výsledky vyhledávání - "Liang, Jiacheng"

Report

Buckle Up: Robustifying LLMs at Every Customization Stage via Data Curation

Autor: Liu, Xiaoqun, Liang, Jiacheng, Tang, Luoxi, You, Chenyu, Ye, Muchao, Xi, Zhaohan

Large language models (LLMs) are extensively adapted for downstream applications through a process known as "customization," with fine-tuning being a common method for integrating domain-specific expertise. However, recent studies have revealed a vul

Externí odkaz: http://arxiv.org/abs/2410.02220

Zobrazit plný text záznamu

Report

Robustifying Safety-Aligned Large Language Models through Clean Data Curation

Autor: Liu, Xiaoqun, Liang, Jiacheng, Ye, Muchao, Xi, Zhaohan

Large language models (LLMs) are vulnerable when trained on datasets containing harmful content, which leads to potential jailbreaking attacks in two scenarios: the integration of harmful texts within crowdsourced data used for pre-training and direc

Externí odkaz: http://arxiv.org/abs/2405.19358

Zobrazit plný text záznamu

Report

Model Extraction Attacks Revisited

Autor: Liang, Jiacheng, Pang, Ren, Li, Changjiang, Wang, Ting

Model extraction (ME) attacks represent one major threat to Machine-Learning-as-a-Service (MLaaS) platforms by ``stealing'' the functionality of confidential machine-learning models through querying black-box APIs. Over seven years have passed since

Externí odkaz: http://arxiv.org/abs/2312.05386

Zobrazit plný text záznamu

Akademický článek

Paleomagnetic constraints on the link between the Comei-Bunbury large igneous province and the Kerguelen mantle plume

Autor: Bian, Weiwei, Yang, Tianshui, Deng, Jin, Ma, Yiming, Peng, Wenxiao, Wang, Suo, Jiao, Xianwei, Ma, Jiahui, Liang, Jiacheng, Jin, Jingjie, Li, Haiyan, Wu, Huaichun, Zhang, Shihong

Publikováno v: In Gondwana Research October 2024 134:30-47

Zobrazit plný text záznamu

Report

OmniLytics: A Blockchain-based Secure Data Market for Decentralized Machine Learning

Autor: Liang, Jiacheng, Li, Songze, Cao, Bochuan, Jiang, Wensi, He, Chaoyang

We propose OmniLytics, a blockchain-based secure data trading marketplace for machine learning applications. Utilizing OmniLytics, many distributed data owners can contribute their private data to collectively train an ML model requested by some mode

Externí odkaz: http://arxiv.org/abs/2107.05252

Zobrazit plný text záznamu