Get started
Author

Donghai Hong

Observed name: Hong, Donghai
Institution: Peking University
Observed institution: Peking University
OpenAlex ID: A5111238960
PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference
2025 · Jiaming Ji, Donghai Hong, et al. · Annual Meeting of the Association for Computational Linguistics (ACL) · 19
PDF
Aligner: Efficient Alignment by Learning to Correct
2024 · Jiaming Ji, Boyuan Chen, et al. · neural information processing systems · 4
PDF
Safe RLHF-V: Safe Reinforcement Learning from Multi-modal Human Feedback
2025 · Jiaming Ji, Xinyu Chen, et al. · neural information processing systems · 3
PDF
PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference
2024 · Ji, Jiaming, Hong, Donghai, et al. · arXiv (Cornell University) · 2
PDF
A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment
2025 · Wang, Kun, Zhang, Guibin, et al. · arXiv (Cornell University) · 1
PDF
Generative RLHF-V: Learning Principles from Multi-modal Human Preference
2025 · Jiayi Zhou, Jiaming Ji, et al. · neural information processing systems · 1
PDF
A Blind Spot in Alignment: Quantifying Biosecurity Risks in Large Language Models
2026 · Quan, Shu, Hao, Tianfang, et al. · arXiv (Cornell University)
PDF
Debate with Images: Detecting Deceptive Behaviors in Multimodal Large Language Models
2025 · Fang, Sitong, Hou, Shiyi, et al. · arXiv (Cornell University)
PDF
AI Deception: Risks, Dynamics, and Controls
2025 · Chen, Boyuan, Fang, Sitong, et al. · arXiv (Cornell University)
PDF
Mitigating Deceptive Alignment via Self-Monitoring
2025 · Ji, Jiaming, Chen, Wenqi, et al. · arXiv (Cornell University)
PDF
InterMT: Multi-Turn Interleaved Preference Alignment with Human Feedback
2025 · Boyuan Chen, Donghai Hong, et al. · neural information processing systems
Align Anything: Training All-Modality Models to Follow Instructions with Language Feedback
2024 · Ji, Jiaming, Zhou, Jiayi, et al. · arXiv (Cornell University)
PDF
Works count: 17
Citation count: 30
H-index: 3
I10-index: 1
Year
Topic
Type