Goodfire 联合创始人兼首席科学家 · Co-founder & Chief Scientist, Goodfire
可解释性公司 Goodfire 联合创始人兼首席科学家。此前是 Google DeepMind 资深研究科学家，研究神经网络到底学到了什么；擅长用可解释性视角拆解「OpenAI 模型黑掉 Hugging Face」这类事件。
Co-founder and chief scientist of interpretability company Goodfire. Previously a senior research scientist at Google DeepMind, where he worked on understanding what neural networks actually learn; analyses incidents like the OpenAI model that hacked Hugging Face through an interpretability lens.
在 AI Podcast 查看 TA 的全部内容 →