The State of Reinforcement Learning for LLM Reasoning

A lot has happened this month, especially with the releases of new flagship models like GPT-4.5 and Llama 4. But you might have noticed that reactions to...