Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics

Hongbin Pei, Jingxin Hai, Yu Li, Huiqi Deng, Denghao Ma +4 more
2/14/2026

No official code found. Best implementation: XJTU-Graph-Intelligence-Lab/two-phasic-for-pseudo-labeling (1 star, last commit 2025-08-22).

Semantic Scholar

Code Implementations(1)

XJTU-Graph-Intelligence-Lab/two-phasic-for-pseudo-labelingCommunity71%

Code of the paper "Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics"

10Python, MakefileMay 22, 20251 years ago

Cite this paper

@article{pei2026nonstationary,
  title  = {Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics},
  author = {Hongbin Pei and Jingxin Hai and Yu Li and Huiqi Deng and Denghao Ma and Jie Ma and Pinghui Wang and Jing Tao and Xiaohong Guan},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:6fa940071f140a749425174df15b6ca5b458670a},
  journal = {ICML 2025 2025}
}

Discussion