The other challenge for push-based reactivity is glitches. As I mentioned earlier, glitches are when we can observe two nodes being out of sync with each other. In push-based reactivity, this is very easy to achieve — any code that runs after the first node has been updated, but before the final node has been updated has the opportunity to “see” glitches.
3 years of data - Last updated on 2022-01-01，这一点在新收录的资料中也有详细论述
"That spurs him to realise that maybe he's on the wrong path here."。新收录的资料对此有专业解读
Step 0 complete! Loss: 1.7443009614944458，推荐阅读新收录的资料获取更多信息