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

Sebastian Raschka's Blog offers insights, tutorials, and research updates on machine learning, deep learning, and artificial intelligence. Covering topics such as neural networks, data science, and Python programming, Sebastian Raschka's Blog provides resources for students, researchers, and practitioners in the field of AI. Developers can learn about  algorithms, research methodologies, and practical applications of machine learning through Raschka's blog posts and publications.

Sebastian Raschka

A comprehensive 2025 review of large language model developments highlights reinforcement learning with verifiable rewards (RLVR) and the GRPO algorithm as the year's dominant training paradigm, following DeepSeek R1's breakthrough. Key trends include inference-time scaling, tool use integration, and architectural efficiency tweaks like mixture-of-experts and linear attention mechanisms. The analysis addresses benchmarking challenges ("benchmaxxing"), discusses practical LLM usage for coding and writing, and examines the shift toward domain-specific models with proprietary data. Predictions for 2026 emphasize RLVR expansion beyond math/code, increased inference optimization, and the emergence of diffusion models for low-latency tasks.

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

2. GRPO, the Research Darling of the Year

3. LLM Architectures: A Fork in the Road?

4. It’s Also The Year of Inference-Scaling and Tool Use

8. Building LLMs and Reasoning Models From Scratch

9. Surprises in 2025 and Predictions for 2026

10. Bonus: A Curated LLM Research Papers List (July to December 2025)

<p>RLVR sounds like an actual solution for further development of LLM models, rather than U/SL (supervised learning)</p>
