人工智能2015 · 分享于 Hacker News
Distilling the Knowledge in a Neural Network
探索《Distilling the Knowledge in a Neural Network》的原文与 Hacker News 社区讨论，查看来源并加入阅读清单。
从论文，走向讨论
看看社区如何讨论这个想法。
来自 Hacker News 的原始讨论。不同的视角、疑问，以及值得继续思考的线索。
最多展示 100 条相关讨论。论文可能出现在评论中，并非原帖投稿内容。
在讨论中发现的引用
按 HN 条目顺序展示最近的 50 条引用来源。
YFede_V ·
AlphaGo beats Lee Sedol 3-0 [video]
566 赞同407 评论
Yxwintermutex ·
How Google Translate squeezes deep learning onto a phone
403 赞同92 评论
Yrwieruch ·
Neural Networks in JavaScript with Deeplearn.js
396 赞同72 评论
Ymortenjorck ·
The Myth of a Superhuman AI
254 赞同357 评论
Ysarthakjain ·
How to use Deep Learning when you have Limited Data
252 赞同50 评论
Ygroar ·
Neurogenesis Deep Learning
197 赞同97 评论
Ypseudolus ·
Distillation makes AI models smaller and cheaper
164 赞同82 评论
YAnon84 ·
Training of Physical Neural Networks
142 赞同46 评论
Ypete_b_condon ·
New Explainable AI Algorithms
111 赞同18 评论
Yjpindar ·
Inside Darpa’s Push to Make Artificial Intelligence Explain Itself
88 赞同47 评论
Yzeveb ·
The crisis of American forensics
74 赞同15 评论
Yprostoalex ·
Scientists discover specific neurons involved in making memories
56 赞同9 评论
Ysirteno ·
Could you train a ChatGPT-beating model for $85,000 and run it in a browser?
8 赞同0 评论
YPaulHoule ·
Efficient Transformer Knowledge Distillation: A Performance Review
7 赞同0 评论
Ytim_sw ·
Distilling the Knowledge in a Neural Network
2 赞同0 评论
Ynabla9 ·
Eliezer Yudkowsky on large language model economics
2 赞同3 评论
Ysdpmas ·
NanoGPT Slowrun: 10x Data Efficiency with Infinite Compute
2 赞同0 评论
Ysoftwaredoug ·
"We have information that Moonshot distilled Fable for the development of K3"
2 赞同1 评论