Deep (Learning) Focus
Subscribe
Sign in
Home
Notes
The Author
Archive
About
Applying Statistics to LLM Evaluations
An overview of useful statistics for building and interpreting LLM evaluations...
READ THE LATEST
Most Popular
View all
Decoder-Only Transformers: The Workhorse of Generative LLMs
Mar 4, 2024
•
Cameron R. Wolfe, Ph.D.
163
15
10
Demystifying Reasoning Models
Feb 18, 2025
•
Cameron R. Wolfe, Ph.D.
277
5
30
Understanding and Using Supervised Fine-Tuning (SFT) for Language Models
Sep 11, 2023
•
Cameron R. Wolfe, Ph.D.
87
5
8
AI Agents from First Principles
Jun 9, 2025
•
Cameron R. Wolfe, Ph.D.
359
25
44
Latest
Top
Discussions
Rubric-Based Rewards for RL
Extending the benefits of large-scale RL training to non-verifiable domains...
Feb 16
107
9
15
Continual Learning with RL for LLMs
Exploring the impressive continual learning capabilities of RL training...
Jan 26
•
Cameron R. Wolfe, Ph.D.
141
14
19
GRPO++: Tricks for Making RL Actually Work
How to go from the vanilla GRPO algorithm to functional RL training at scale...
Jan 5
•
Cameron R. Wolfe, Ph.D.
128
10
18
Olmo 3 and the Open LLM Renaissance
Fully-open artifacts with the potential to make LLM research a reality for anyone...
Dec 15, 2025
•
Cameron R. Wolfe, Ph.D.
81
7
14
Group Relative Policy Optimization (GRPO)
How the algorithm that teaches LLMs to reason actually works...
Nov 24, 2025
•
Cameron R. Wolfe, Ph.D.
106
10
14
PPO for LLMs: A Guide for Normal People
Understanding the complex RL algorithm that gave us modern LLMs…
Oct 27, 2025
•
Cameron R. Wolfe, Ph.D.
164
11
14
REINFORCE: Easy Online RL for LLMs
How to get the benefits of online RL without the complexity of PPO...
Sep 29, 2025
•
Cameron R. Wolfe, Ph.D.
99
11
6
See all
Deep (Learning) Focus
I contextualize and explain important topics in AI research.
Subscribe
Recommendations
View all 13
Javarevisited Newsletter
javinpaul
The Founders Corner®
Ruben Dominguez
AI Newsletter
elvis
Ahead of AI
Sebastian Raschka, PhD
LLM Watch
Pascal Biese
Deep (Learning) Focus
Subscribe
About
Archive
Recommendations
Sitemap
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts