The State Of LLMs 2025: Progress, Problems, and Predictions

A 2025 review of large language models, from DeepSeek R1 and RLVR to inference-time scaling, benchmarks, architectures, and predictions for 2026.

Dec 30, 2025 • Sebastian Raschka, PhD

442

35

49

LLM Research Papers: The 2025 List (July to December)

In June, I shared a bonus article with my curated and bookmarked research paper lists to the paid subscribers who make this Substack possible.

Dec 30, 2025 • Sebastian Raschka, PhD

31

1

From DeepSeek V3 to V3.2: Architecture, Sparse Attention, and RL Updates

Understanding How DeepSeek's Flagship Open-Weight Models Evolved

Dec 3, 2025 • Sebastian Raschka, PhD

245

12

28

November 2025

Beyond Standard LLMs

Linear Attention Hybrids, Text Diffusion, Code World Models, and Small Recursive Transformers

Nov 4, 2025 • Sebastian Raschka, PhD

352

25

36

October 2025

Understanding the 4 Main Approaches to LLM Evaluation (From Scratch)

Multiple-Choice Benchmarks, Verifiers, Leaderboards, and LLM Judges with Code Examples

Oct 5, 2025 • Sebastian Raschka, PhD

353

26

33

September 2025

Understanding and Implementing Qwen3 From Scratch

A Detailed Look at One of the Leading Open-Source LLMs

Sep 6, 2025 • Sebastian Raschka, PhD

114

6

10

August 2025

From GPT-2 to gpt-oss: Analyzing the Architectural Advances

And How They Stack Up Against Qwen3

Aug 9, 2025 • Sebastian Raschka, PhD

616

47

55

July 2025

The Big LLM Architecture Comparison

From DeepSeek-V3 to Kimi K2: A Look At Modern LLM Architecture Design

Jul 19, 2025 • Sebastian Raschka, PhD

1,700

85

155

LLM Research Papers: The 2025 List (January to June)

A topic-organized collection of 200+ LLM research papers from 2025

Jul 1, 2025 • Sebastian Raschka, PhD

95

5

9

June 2025

Understanding and Coding the KV Cache in LLMs from Scratch

KV caches are one of the most critical techniques for efficient inference in LLMs in production.

Jun 17, 2025 • Sebastian Raschka, PhD

435

37

34

May 2025

Coding LLMs from the Ground Up: A Complete Course

Why build LLMs from scratch? It's probably the best and most efficient way to learn how LLMs really work. Plus, many readers have told me they had a lot…

May 10, 2025 • Sebastian Raschka, PhD

251

4

18

April 2025

The State of Reinforcement Learning for LLM Reasoning

Understanding GRPO and New Insights from Reasoning Model Papers

Apr 19, 2025 • Sebastian Raschka, PhD

486

33

39