Stealthy and Persistent Unalignment on Large Language Models via Backdoor Injections von Cao, Yuanpu, Cao, Bochuan, Chen, Jinghui

Volltext bestellen

Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM von Cao, Bochuan, Cao, Yuanpu, Lin, Lu, Chen, Jinghui

Volltext bestellen

Data Free Backdoor Attacks von Cao, Bochuan, Jia, Jinyuan, Hu, Chuxuan, Guo, Wenbo, Xiang, Zhen, Chen, Jinghui, Li, Bo, Song, Dawn

Volltext bestellen

On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept von Liu, Guangliang, Mao, Haitao, Cao, Bochuan, Xue, Zhiyu, Zhang, Xitong, Wang, Rongrong, Tang, Jiliang, Johnson, Kristen

Volltext bestellen

On the Difficulty of Defending Contrastive Learning against Backdoor Attacks von Li, Changjiang, Pang, Ren, Cao, Bochuan, Xi, Zhaohan, Chen, Jinghui, Ji, Shouling, Wang, Ting

Volltext bestellen

IMPRESS: Evaluating the Resilience of Imperceptible Perturbations Against Unauthorized Data Usage in Diffusion-Based Generative AI von Cao, Bochuan, Li, Changjiang, Wang, Ting, Jia, Jinyuan, Li, Bo, Chen, Jinghui

Volltext bestellen

On the Safety of Open-Sourced Large Language Models: Does Alignment Really Prevent Them From Being Misused? von Zhang, Hangfan, Guo, Zhimeng, Zhu, Huaisheng, Cao, Bochuan, Lin, Lu, Jia, Jinyuan, Chen, Jinghui, Wu, Dinghao

Volltext bestellen

OmniLytics: A Blockchain-based Secure Data Market for Decentralized Machine Learning von Liang, Jiacheng, Li, Songze, Cao, Bochuan, Jiang, Wensi, He, Chaoyang

Volltext bestellen

Wild-Time: A Benchmark of in-the-Wild Distribution Shift over Time von Yao, Huaxiu, Choi, Caroline, Cao, Bochuan, Lee, Yoonho, Koh, Pang Wei, Finn, Chelsea

Volltext bestellen

Defending Against Alignment-Breaking Attacks via Robustly Aligned LLM von Cao, Bochuan, Cao, Yuanpu, Lu, Lin, Chen, Jinghui

Volltext

On the Intrinsic Self-Correction Capability of LLMs: Uncertainty and Latent Concept von Liu, Guangliang, Mao, Haitao, Cao, Bochuan, Xue, Zhiyu, Zhang, Xitong, Wang, Rongrong, Tang, Jiliang, Johnson, Kristen

Volltext

WordGame: Efficient & Effective LLM Jailbreak via Simultaneous Obfuscation in Query and Response von Zhang, Tianrong, Cao, Bochuan, Cao, Yuanpu, Lu, Lin, Mitra, Prasenjit, Chen, Jinghui

Volltext

Watch the Watcher! Backdoor Attacks on Security-Enhancing Diffusion Models von Li, Changjiang, Pang, Ren, Cao, Bochuan, Chen, Jinghui, Ma, Fenglong, Ji, Shouling, Wang, Ting

Volltext

On the Difficulty of Defending Contrastive Learning against Backdoor Attacks von Li, Changjiang, Pang, Ren, Cao, Bochuan, Zhaohan Xi, Chen, Jinghui, Ji, Shouling, Wang, Ting

Volltext

On the Safety of Open-Sourced Large Language Models: Does Alignment Really Prevent Them From Being Misused? von Zhang, Hangfan, Guo, Zhimeng, Zhu, Huaisheng, Cao, Bochuan, Lu, Lin, Jia, Jinyuan, Chen, Jinghui, Wu, Dinghao

Volltext

OmniLytics: A Blockchain-based Secure Data Market for Decentralized Machine Learning von Liang, Jiacheng, Li, Songze, Cao, Bochuan, Jiang, Wensi, He, Chaoyang

Volltext

Pumpkin Leaf Extract Crop Waste as a New Degradable and Environmentally Friendly Corrosion Inhibitor von Ren, Haiqin, Liu, Yan, Gong, Zhili, Tan, Bochuan, Deng, Hongda, Xiong, Junle, Shao, Peng, Dai, Qingwei, Cao, Jiangtao, Marzouki, Riadh

Volltext

Impact of upgrading wastewater treatment plant on the removal of typical methyl, oxygenated, chlorinated and parent polycyclic aromatic hydrocarbons von Qiao, Meng, Cao, Wei, Liu, Bochuan, Bai, Yaohui, Qi, Weixiao, Zhao, Xu, Qu, Jiuhui

Volltext

Formation of smooth anodic nanoporous iron oxide film for enhancing photocathodic protection on plain carbon steel von Deng, Hongda, Gou, Xiantao, Chen, Qilin, Tan, Bochuan, Cao, Xianlong

Volltext

Comparison of flavins and a conjugated oligoelectrolyte in stimulating extracellular electron transport from Shewanella oneidensis MR-1 von Wang, Victor Bochuan, Kirchhofer, Nathan D., Chen, Xiaofen, Tan, Melissa Yuan Li, Sivakumar, Krishnakumar, Cao, Bin, Zhang, Qichun, Kjelleberg, Staffan, Bazan, Guillermo C., Loo, Say Chye Joachim, Marsili, Enrico

Volltext