AlphaToken: Decoupling Adaptation and Stability for Path-Aware Response Token Valuation in LLM Post-Training

@misc{qing2026alphatoken,
  title = {AlphaToken: Decoupling Adaptation and Stability for Path-Aware Response Token Valuation in LLM Post-Training},
  author = {Liu Qing and Ou Wu and Yi Du},
  year = {2026},
  eprint = {2606.01635},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2606.01635},
}