Learning Neural Trans-Dimensional Random Field Language Models with Noise-Contrastive Estimation

方法与贡献

对参考分布进行指数倾斜，并用噪声对比估计训练神经随机场语言模型。

2020 年论文明确讨论的更早残差形式与 NCE 训练方法，其评估以语音识别重评分为中心。

结合主题阅读

基于能量的文本生成：残差模型、重采样与扩散 →

引用

Bin Wang, Zhijian Ou. Learning Neural Trans-Dimensional Random Field Language Models with Noise-Contrastive Estimation. ICASSP 2018.

@inproceedings{neuraltrfnce,
  title = {{Learning Neural Trans-Dimensional Random Field Language Models with Noise-Contrastive Estimation}},
  author = {Wang, Bin and Ou, Zhijian},
  year = {2018},
  booktitle = {ICASSP 2018},
  url = {https://arxiv.org/abs/1710.10739},
  eprint = {1710.10739},
  archivePrefix = {arXiv}
}