titleInherited, Not Invented: Tracing the Provenance of Wait in Reasoning Models
authorsZhizhuo Kou, Sizai Hou, Long Chen, Jian Yang, Junyu Luo, Jingshu Peng, Wenkai Zhang, Sirui Han, Yike Guo
date2026-10-22
venuearXiv (preprint, not peer reviewed)
idarXiv:2610.30882
affiliationHUAWEI Noah's Ark Lab
upvotes16 (Hugging Face Daily Papers, 2026-11-01)
linksabs pdf hf code
abstractReinforcement 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' training histories, reading it with label-free indices of frequency, position, form and usage and an injected-error test, all computed from sampled outputs. We find that zero-RL almost never coins the marker: on bases that lack the word, RL raises MATH-500 accuracy substantially while {wait} stays near absent. It arrives with training data: mid-training in OLMo-3, and fine-tuning on reasoning traces installs it in most models whose base lacks it. Where the base already uses it, both released RL-Zero recipes (OLMo-3, MiMo) multiply it, whereas KL-free GRPO from the same base does not; OLMo-3's recipe wears down its form and makes it a more sensitive but more liberal alarm; RL after supervised fine-tuning barely changes it, and our GRPO recipe erases a marker seeded by fine-tuning. Finally, we show that marker frequency alone does not indicate RL-induced reasoning and caution against LLM-as-a-judge revision metrics, as automated judges can reverse human-labelled trends.

cite

@misc{kou2026inherited,
  title={Inherited, Not Invented: Tracing the Provenance of Wait in Reasoning Models},
  author={Zhizhuo Kou and Sizai Hou and Long Chen and Jian Yang and Junyu Luo and Jingshu Peng and Wenkai Zhang and Sirui Han and Yike Guo},
  year={2026},
  eprint={2610.30882},
  archivePrefix={arXiv},
  url={https://arxiv.org/abs/2610.30882}
}