01
Sebastian Raschka · 2025 · Ahead of AI
Why?
About 200 papers sorted by topic rather than date, which is what makes it usable as a map instead of a feed: reasoning models, RL for LLMs, inference-time scaling, efficient architectures, diffusion LMs, multimodal, and data. Read the abstracts, pick one paper per section, and stop. The July to December half continues the same categories.