tweetindex
EN

Ethan He

@EthanHe_42 · Palo Alto, CA · joined 15 Mar 2015

ex world model lead @xAI | ex @Nvidia @Meta | 30+ papers, 9k citations | talk about AI, LLM, video generation, multimodal, AGI

37 212Followers
704Following
1 319Posts total
251.8KViews on collected posts

Latest posts

@EthanHe_42 Are you using Super @Grok? https://t.co/NsIWS9WXFo 9 views · 0 likes · 0 reposts · 1 replies Open on X →
@EthanHe_42 @latentspacepod Ethan you are amazing 🙌🏽 2K views · 5 likes · 0 reposts · 0 replies Open on X →
@latentspacepod Apple podcast: https://t.co/eMH1FzNSjk Spotify: https://t.co/UQKNOldLY0 transcript on Substack: https://t.co/jvxwiCk92c 5K views · 14 likes · 2 reposts · 1 replies Open on X →
In @latentspacepod podcast, I shared my view on video generation, world models, LLMs, agents, continual learning and where the next frontier is. 1. Video models get most of their intelligence from language, not from video data. 2. Idea-to-code is fast now. The bottleneck is back 141.7K views · 380 likes · 40 reposts · 25 replies Open on X →
https://t.co/xJAJLZhIWK 4.3K views · 48 likes · 4 reposts · 0 replies Open on X →
"You can outsource thinking, but not understanding." I still find writing toy code one of the best ways to build real understanding. It catches the nuances that skimming code and explanations lets you skip. So I wrote nanoRL (nanoGPT, but for post-training). SFT, DPO, GRPO, ht 44.2K views · 771 likes · 53 reposts · 22 replies Open on X →
We applied AlphaGo's algorithm to video generation. Long video generation often breaks after a few extensions. We use MCTS to evaluate multiple continuations with look-ahead rollouts and backpropagated rewards. It produces long video while maintaining comparable visual https:// 31.3K views · 418 likes · 29 reposts · 16 replies Open on X →
Humans are the only species that spends 20% of their energy on the brain. Most animals are well below 10%. The brain is "useless" in the direct survival sense. It doesn't catch prey, doesn't run, doesn't fight. Spending 20% of precious calories on being smarter has to help https 18.3K views · 237 likes · 19 reposts · 24 replies Open on X →
NVIDIA-NeMo/DFM got archived this week. last summer a few of us bootstrapped it with 100k+ lines of code, scaling diffusion models to 100B+ with full 5D parallelism. that was months of painful grind. might be a weekend in this vibe coding era? on to the next one. https://t.co/UjX 5K views · 45 likes · 3 reposts · 2 replies Open on X →

Against accounts of the same size

1 posts from the last 90 days, next to the 10K–100K follower range. reaches fewer people than peers, but engages them much harder.

Median views9this account944median for 10K–100K
Reach, %0.02%this account3.56%median for 10K–100K
Engagement, %11.11%this account1.96%median for 10K–100K
MetricThis accountMedian for 10K–100KRatio
Median views per post99440.01×
Reach (views ÷ followers)0.02%3.56%0.01×
Engagement rate11.11%1.96%5.67×

Others in this range →   Compare with another account →   How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

5K22 May
18.3K26 May
31.3K27 May
44.2K1 Jun
4.3K
141.7K
5K
2K
928 Aug

Last 9 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

1.00%22 May
1.54%26 May
1.49%27 May
1.93%1 Jun
1.21%
0.32%
0.34%
0.24%
11.11%28 Aug

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.96%.

What the audience does

Likes57.6%1 918 in total
Reposts4.5%150 in total
Replies2.7%91 in total
Quotes0.8%25 in total
Bookmarks34.4%1 145 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Similar accounts