齐科·科尔特 Zico KolterZico Kolter
Carnegie Mellon / OpenAI Board
CMU 教授 · OpenAI 董事 · AI 安全Professor at CMU · OpenAI Board · AI Safety
safetydeep-learning
美国计算机科学家，卡内基梅隆大学教授兼机器学习系主任。研究涵盖对抗鲁棒性、深度均衡模型，以及把优化作为神经网络的一层（OptNet）。2023 年其团队发表了被广泛引用的工作，展示可通用、可迁移的对抗攻击能越狱已对齐的大模型。2024 年加入 OpenAI 董事会并担任其安全与安保委员会主席，还联合创办 AI 安全创业公司 Gray Swan。在让前沿模型更稳健、更安全的方向上有重要工作。
American computer scientist, professor and head of the Machine Learning Department at Carnegie Mellon University. His research spans adversarial robustness, deep equilibrium models, and optimization as a layer inside neural networks (OptNet). In 2023 his group published work on universal, transferable adversarial attacks that jailbreak aligned LLMs. He joined the OpenAI board of directors in 2024 and chairs its Safety and Security Committee, and co-founded the AI-security startup Gray Swan. A leading academic voice on making frontier models robust and safe.
在关系图谱中查看 齐科·科尔特 →View Zico Kolter in the graph →
听 齐科·科尔特 的 AI 播客 →Listen to Zico Kolter on AI Podcast →
时间线Timeline
- 2019 Deep Equilibrium Models
- 2023 Universal and Transferable Adversarial Attacks on Aligned LLMs