A new analysis argues that the word "wait" in reasoning models' chains of thought is inherited from training data rather than taught by reinforcement learning, noting that large RL accuracy gains occur without the word. The work traces the token through the training corpora of open models to separate what RL discovers from what a model inherits.
Reinforcement learning (RL) is credited with teaching reasoning models to write {wait} before reconsidering a step, yet large RL accuracy gains do not require the word. This view can confuse what RL discovers with what a model inherits from its training data. We trace the word through open models' t
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.