←新闻详情

谢赛宁更新于 10月14日 22:32

谢赛宁推荐 RAEv2:强表征编码器对像素解码器至关重要

check out RAEv2 led by Jas. through extensive exps, we found some really intriguing behaviors showing why strong representation encoders are key for pixel decod

AI 导读

看看 Jas 主导的 RAEv2。通过大量实验,我们发现了一些非常有趣的行为,表明为什么强表征编码器对像素解码器至关重要。剧透:关键不在于爬 FID 的坡;像 ep@fid-k/fdr^k 这样的新指标表明还有更多值得探索的空间!

引用

Jaskirat Singh

@1jaskiratsingh

In Oct last year, Representation Autoencoders provided an elegant solution to unified tokenization for understanding and generation. Today we make them a bit more simple. a bit more general. Result: >10x faster convergence, better reconstruction, better generation. And yes we test them on T2I and world models :) Introducing RAEv2

查看原帖

正文

当前显示原文,中文全文译文尚未就绪。

check out RAEv2 led by Jas. through extensive exps, we found some really intriguing behaviors showing why strong representation encoders are key for pixel decoders. spoiler: it’s not about hillclimbing fid; new metrics like ep@fid-k/fdr^k show there’s a lot more left to explore!

来源

谢赛宁