Get started
Author

Boyi Wei

Observed name: Wei, Boyi
OpenAlex ID: A5101332910
A benchmark of expert-level academic questions to assess AI capabilities
2026 · Long Phan, Alice Gatti, et al. · Nature · 65
PDF
SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal
2024 · Xie, Tinghao, Qi, Xiangyu, et al. · arXiv (Cornell University) · 7
PDF
AI Risk Management Should Incorporate Both Safety and Security
2024 · Qi, Xiangyu, Huang, Yangsibo, et al. · arXiv (Cornell University) · 3
PDF
An Adversarial Perspective on Machine Unlearning for AI Safety
2024 · Łucki, Jakub, Wei, Boyi, et al. · arXiv (Cornell University) · 2
PDF
Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications
2024 · Wei, Boyi, Huang, Kaixuan, et al. · arXiv (Cornell University) · 2
PDF
Understanding and Reducing Misalignment in Persuasion Optimization
2026 · Wei, Boyi, Desai, Ruta, et al. · arXiv (Cornell University)
PDF
Large Language Models Generate Harmful Responses Using a Distinct Mechanism, Shared Across Harm Types
2026 · Orgad, Hadas, Wei, Boyi, et al. · arXiv (Cornell University)
PDF
A benchmark of expert-level academic questions to assess AI capabilities
2026 · Center for AI Safety, Phan, Long, et al. · RWTH Publications (RWTH Aachen)
Dynamic Risk Assessments for Offensive Cybersecurity Agents
2025 · Boyi Wei, Benedikt Stroebl, et al. · neural information processing systems
PDF
Works count: 10
Citation count: 80
H-index: 3
I10-index: 1
Year
Topic
Type