迟明远 · Walker
I do LLM pretraining — currently at ████████, previously at MiniMax, and before that Tencent TEG as an intern. Background in Computer Science (HUST, graph neural networks) and Computational Science & Engineering (MSc, ETH Zürich; numerical PDE, robotics simulation, sparse attention for LLMs; exchange at RWTH Aachen). My research interests are LLMs, multimodal LLMs, GPU kernels and DSL compilers, and numerical PDE simulation. The Three-Body Problem is my bible.
我做 LLM 预训练,目前在 ████████, 之前在 MiniMax,再之前在腾讯 TEG 实习。本科在华科计算机做图神经网络, 硕士在苏黎世联邦理工的计算科学工程,研究 PDE 数值仿真、机器人仿真、 以及 LLM 的 sparse attention(其间去亚琛工大交换过一学期)。 我的科研兴趣是大语言模型、多模态大模型、GPU kernel / DSL 编译器,以及 PDE 数值仿真, 《三体》就是我的圣经。