诺姆·布朗 Noam BrownNoam Brown
OpenAI
推理研究 · 扑克 AI → o1Reasoning · Poker AI → o1
rldeep-learning
CMU 博士。创造了 Libratus（2017）和 Pluribus（2019），首个在单挑和多人无限注德州扑克中击败顶级职业选手的 AI 系统。先在 Meta AI Research (FAIR) 工作。2023年被 OpenAI 招募，成为 o1 推理模型的关键贡献者，将博弈论搜索和规划方法应用于 LLM 推理。他的「测试时计算」（让模型思考更久）工作是 OpenAI 推理突破的核心。
CMU PhD. Created Libratus (2017) and Pluribus (2019), the first AI systems to beat top poker professionals at both heads-up and multiplayer no-limit Texas Hold'em. Joined Meta AI Research (FAIR). Recruited by OpenAI in 2023, became a key contributor to the o1 reasoning model, applying game-theoretic search and planning to LLM inference. His work on 'test-time compute' (letting models think longer) is central to OpenAI's reasoning breakthrough.
在关系图谱中查看 诺姆·布朗 →View Noam Brown in the graph →
时间线Timeline
- 2017 Superhuman AI for heads-up no-limit poker: Libratus (Science)
- 2019 Superhuman AI for multiplayer poker (Science)
- 2024 o1 reasoning (key contributor)