<aside>
"A seeker of truth. An explorer of reality. A student of life."
</aside>
Hi, I am 帅先 Alex 👋. Belows are some of my posts.
To know more about me, visit my website here: 👉 link
[2601] online learning 的新论文：TTT-Discover
[2601] **关于大模型的下一个范式的思考之 online learning** 🌟
[2507] **Grok-4 Deep Dive - RL 算力和数据拆解** 🌟
[2506] **一文弄懂 RL：REINFORCE、PPO、GRPO 到底在做啥** 🌟
[2506] 再读 The Bitter Lesson 和 The Era of Experience
[2505] Deepseek’s insight on V3 and infra 论文拆解（ISCA‘25）
[2505] Linear Attention vs. Transformer: Q&A
[2504] 各种 Normalization 如何影响 LLM？
[2503] ByteScale：字节跳动的万卡长序列训练方案
[2512] **Semi 101 (Part 1)：制造 ，Semi 101 (Part 2)：硬件 & 系统**
[2511] **AI Coding 行业 Deep Dive** 🌟
[2510] AI infra 建设狂潮下，电力行业如何看？
[2509] **大模型 serving 经济学：「TPS、token 价格、激活参数」三角** 🌟
[2508] 一年后的今天，我们能够回答红杉的 AI’s $600B Question 了吗？
[2506] 博通 Broadcom Deep Dive - 基本面分析
[2403] 从 GTC 2024 看 Blackwell 到底强在哪里?
Some of my early research efforts, summarized here: My Research Overview
Previously, I maintained a knowledge base on GitHub. Some highlighted notes: