Neel Nanda ✓

@NeelNanda5

Mechanistic Interpretability lead DeepMind. Formerly @AnthropicAI, independent. In this to reduce AI X-risk. Neural networks can be understood, let's go do it!

44 312Followers

122Following

5 359Posts total

332.6KViews on collected posts

Posts recentes

Neel Nanda ✓@NeelNanda54 Sep

@NeelNanda5 What if we project every loop iteration output onto logits or through a J-lens? 59 views · 0 likes · 0 reposts · 0 replies Open on X →

Neel Nanda ✓@NeelNanda54 Sep

@NeelNanda5 I saw a paper today about triple extremes for "looped transformers". I honestly don't understand why they think something's better than quadratics. We can form looped transformers anytime in inference, while preserving observability, simply chainning CoT and Meta rea 76 views · 0 likes · 0 reposts · 0 replies Open on X →

Neel Nanda ✓@NeelNanda53 Sep

@NeelNanda5 Would you never trade interpretability for performance? Evidence seems to suggest market prefers the latter over the former, perhaps until it's too late to change the preference 😅 155 views · 0 likes · 0 reposts · 1 replies Open on X →

Neel Nanda ✓@NeelNanda53 Sep

@NeelNanda5 agreed, we need both and losing cot monitoring would be bad. we’ve also been finding that cot and interp monitors catch different stuff, so they stack well 849 views · 17 likes · 0 reposts · 0 replies Open on X →

Neel Nanda ✓@NeelNanda53 Sep

This take is partially inspired by discourse about the OpenAI looped transformer, but worth saying in general. I'm not too worried about a model with 2x the number of effective layers of GPT-4, but the risk of a slippery slope to a large number loops seems super bad. 5.1K views · 152 likes · 5 reposts · 3 replies Open on X →

Neel Nanda ✓@NeelNanda53 Sep

A concerningly common take seems to be that keeping Chain of Thought monitorable doesn't matter because interpretability will save us, or it's already useless This is total bullshit. CoT is our best current tool for safety & interpretability, losing it would be a major tragedy 57.2K views · 1.4K likes · 127 reposts · 41 replies Open on X →

Neel Nanda ✓@NeelNanda51 Aug

@NeelNanda5 Prepare to do 16 rounds of interview 1.2K views · 5 likes · 0 reposts · 1 replies Open on X →

Neel Nanda ✓@NeelNanda531 Jul

Apply here: UK: https://t.co/WmcxOukjv9 US: https://t.co/f4yyBY4rs2 Note the guidance in our hiring post on how to apply! Real reasons >> corporate waffle https://t.co/T7UXtAJDIN 27.1K views · 61 likes · 3 reposts · 4 replies Open on X →

Neel Nanda ✓@NeelNanda531 Jul

We're hiring for many roles, in all areas we work on, including: helping align production Gemini, monitoring Gemini deployments for harm/misalignment, evaluating risks from Gemini, methods to align ASI, model organisms, and more Read more on our new blog: https://t.co/YQsAZnWqKU 5.9K views · 38 likes · 1 reposts · 3 replies Open on X →

Neel Nanda ✓@NeelNanda531 Jul

More than a specific role, we're looking for excellence: fantastic researchers, fantastic engineers, great taste, etc All coding interviews allow agents! Read more about the role, what we're looking for, the impact case for applying, FAQs, and more: https://t.co/cqKJj0H82W http 6K views · 66 likes · 1 reposts · 3 replies Open on X →

Neel Nanda ✓@NeelNanda531 Jul

GDM AGI Safety is hiring! Open roles on all subteams, Lon/Bay/more There's lots to do to reduce risks from AGI at Google, but we're bottlenecked on people. If you want to help, please apply! I really love this team and my excellent colleagues. I'm excited to hire more! https:// 162.3K views · 554 likes · 32 reposts · 13 replies Open on X →

Neel Nanda ✓@NeelNanda530 Jul

When looking for RS positions after my PhD I applied to FAIR, Google Brain and DeepMind. FAIR and Brain had 5-6 itws + research talk each. I had 16 itws with DeepMind (the notorious 2hr quizz + 1 coding + 6 remote research itws then 8 onsite research itws and talk). Just looked 66.6K views · 321 likes · 16 reposts · 6 replies Open on X →

Em comparação com contas do mesmo porte

6 posts dos últimos 90 dias, ao lado da faixa de 10K–100K seguidores. alcança menos gente do que contas do mesmo porte.

Mediana de visualizações502esta conta924mediana para 10K–100K

Alcance, %1.13%esta conta3.62%mediana para 10K–100K

Engajamento, %1.32%esta conta1.52%mediana para 10K–100K

Métrica	Esta conta	Mediana para 10K–100K	Proporção
Mediana de visualizações por post	502	924	0.54×
Alcance (visualizações ÷ seguidores)	1.13%	3.62%	0.31×
Taxa de engajamento	1.32%	1.52%	0.87×

Outras contas desta faixa → Comparar com outra conta → Como estas referências são construídas →

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

66.6K30 Jul

162.3K31 Jul

5.9K

27.1K

1.2K1 Aug

57.2K3 Sep

5.1K

849

155

764 Sep

Last 12 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.52%30 Jul

0.37%31 Jul

1.16%

0.72%

0.25%

0.50%1 Aug

2.71%3 Sep

3.13%

2.00%

0.65%

0.00%4 Sep

0.00%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes72.1%2 587 in total

Reposts5.2%185 in total

Replies2.1%75 in total

Quotes0.6%22 in total

Bookmarks20.0%718 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.

Contas semelhantes

Frequently asked questions about @NeelNanda5

How many followers does @NeelNanda5 have?

44 312 followers.

How many posts has @NeelNanda5 published?

5 359 posts.

How do posts by @NeelNanda5 perform?

Collected posts: 12. Views: 332 559. Likes: 2 587.