保留正文与补采附件。源页面当前未重新核验；具体限制见交付清单。
SubscribeSign in Home Notes The Author Archive About
LatestTopDiscussions
Notes on NVIDIA Nemotron
A broad coverage of the most practically useful series of open LLM tech reports...
Sep 28 • Cameron R. Wolfe, Ph.D.
47
9
Reinforcement Learning for LLMs: The Complete Guide
Tracing the evolution of RL from first principles to the frontier of modern AI research...
Aug 24 • Cameron R. Wolfe, Ph.D.
345
10
43
Notes on Midtraining
Creating better specialized LLMs with midtraining and continual pretraining...
Aug 10 • Cameron R. Wolfe, Ph.D.
73
8
Agentic World Models
Creating better language agents by teaching them to model their environment...
Jul 20 • Cameron R. Wolfe, Ph.D.
128
9
19
Agentic RL: Frameworks and Best Practices
How LLMs are trained to handle long horizon tasks in complex environments...
Jun 22 • Cameron R. Wolfe, Ph.D.
194
14
20
Agent Evaluation: A Detailed Guide
Best practices and common patterns for effectively evaluating AI agents...
May 18 • Cameron R. Wolfe, Ph.D.
338
26
53
RL Scaling Laws for LLMs
How scaling laws have evolved from pretraining to reinforcement learning...
Apr 20 • Cameron R. Wolfe, Ph.D.
144
3
17
The Anatomy of an LLM Benchmark
Common patterns used to create the most effective LLM evaluation datasets...
Mar 30 • Cameron R. Wolfe, Ph.D.
118
4
17
Applying Statistics to LLM Evaluations
An overview of useful statistics for building and interpreting LLM evaluations...
Mar 9 • Cameron R. Wolfe, Ph.D.
148
9
15
Rubric-Based Rewards for RL
Extending the benefits of large-scale RL training to non-verifiable domains...
Feb 16
134
11
18
Continual Learning with RL for LLMs
Exploring the impressive continual learning capabilities of RL training...
Jan 26 • Cameron R. Wolfe, Ph.D.
160
15
20
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.
167
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.
85
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.
136
11
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.
185
13
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.
106
11
6
Online versus Offline RL for LLMs
A deep dive into the online-offline performance gap in LLM alignment...
Sep 8, 2025 • Cameron R. Wolfe, Ph.D.
93
4
10
GPT-oss from the Ground Up
Everything you should know about OpenAI's new open-weight language models...
Aug 18, 2025 • Cameron R. Wolfe, Ph.D.
105
14
15
Direct Preference Optimization (DPO)
How to align LLMs with limited hardware and minimal complexity...
Jul 28, 2025 • Cameron R. Wolfe, Ph.D.
142
20
11
Reward Models
Modeling human preferences for LLMs in the age of reasoning models...
Jun 30, 2025 • Cameron R. Wolfe, Ph.D.
145
13
20
AI Agents from First Principles
Understanding AI agents by building upon the most basic concepts of LLMs...
Jun 9, 2025 • Cameron R. Wolfe, Ph.D.
381
26
46
A Guide for Debugging LLM Training Data
Data-centric techniques and tools that anyone should use when training an LLM...
May 19, 2025 • Cameron R. Wolfe, Ph.D.
89
3
13
Llama 4: The Challenges of Creating a Frontier-Level LLM
The full story behind Llama 4 and Meta's huge pivot in research strategy...
Apr 28, 2025 • Cameron R. Wolfe, Ph.D.
83
2
11
Vision Large Language Models (vLLMs)
Teaching LLMs to understand images and videos in addition to text...
Mar 31, 2025 • Cameron R. Wolfe, Ph.D.
156
15
14
© 2026 Cameron R. Wolfe · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture