Two Live Book Club Q&As on September 3
I’ve been getting lots of nice feedback on my new book, Build a Reasoning Model (From Scratch). Thank you!
I am trying hard to answer all questions readers may have about the book and related topics, whether they come through social media, GitHub, or email. However, I may have missed a few things here and there.
I am also looking forward to doing two live Q&As with readers on September 3!
The first Q&A is with Sophia Yang’s The AI Book Club at 10 am CT.
A bit later that same day, at 2 pm CT, I’ll join Sage Elliott’s AI Book Club for a second live Q&A.
Some readers were also curious about Build a Large Language Model (From Scratch), so we may cover how the two books connect. In short, Build a Large Language Model (From Scratch) builds and pre-trains an LLM from the ground up. The newer book is basically a sequel that continues with inference-time scaling and reasoning-focused post-training.
I’m looking forward to hearing which parts readers found most interesting and what questions they have. If you are reading either book, or are simply curious about reasoning models, you are very welcome to join us!
The book clubs and sign-ups are free!
- Sophia Yang’s The AI Book Club (https://luma.com/djttn0ah)
- Sage Elliott’s AI Book Club (https://luma.com/zjg8jx5e)

Read Next
Pre-Training Without Learning Rate Decay
A quick comparison of LLM pre-training recipes using cosine decay, warmup-stable-decay, or constant learning rates with checkpoint averaging.
From Full to Hybrid Attention in LLMs
How open-weight LLMs are shifting from full attention to hybrid and sparse designs, and what that means for long-context compute and memory.
Open October LLM Architecture Roundup
A roundup of ten open-weight LLM architecture releases from October 2026, covering attention mechanisms, hybrid layers, and MoE designs.