郑之杰的个人博客
欢迎光临
Language Models are Few-Shot Learners
GPT3:语言模型是少样本学习器.
Deep Ensembles: A Loss Landscape Perspective
探讨深度集成学习的损失曲面.
When BERT Plays the Lottery, All Tickets Are Winning
讨论BERT剪枝中的“彩票假设”.
Deep image reconstruction from human brain activity
从人类大脑活动中重构图像.
Investigating Human Priors for Playing Video Games
探究电子游戏的人类先验知识.
Meta-Learning with Implicit Gradients
iMAML:使用L2正则化增加内部梯度更新次数.