Sander Dieleman ✓

@sedielem · London, England · joined 02 Dec 2014

Research Scientist at Google DeepMind (WaveNet, Nano Banana, Gemini Omni). I tweet about ML, music, generative models (personal account).

70 599Followers

2 045Following

2 814Posts total

271.4KViews on collected posts

Derniers posts

Sander Dieleman ✓@sedielem14 Sep

@sedielem @doomie I think the position is not posted for Zürich on the website 2.8K views · 5 likes · 0 reposts · 2 replies Open on X →

Sander Dieleman ✓@sedielem14 Sep

@sedielem That looks exciting, thanks for posting. Someone already asked about a referral, but I guess it can't hurt to have one? 1.2K views · 0 likes · 0 reposts · 1 replies Open on X →

Sander Dieleman ✓@sedielem14 Sep

@sedielem Is a referral needed? 1.4K views · 1 likes · 0 reposts · 1 replies Open on X →

Sander Dieleman ✓@sedielem14 Sep

We are hiring research scientists and engineers🧑‍🔬 on the Gemini Omni team! Both in the US (Mountain View / SF) and Europe (London / Zürich). Apply here: https://t.co/8bNvJWQAAL https://t.co/sqHuJRoGbF

72.1K views · 587 likes · 54 reposts · 9 replies Open on X →

Sander Dieleman ✓@sedielem8 Sep

10 years today since we unveiled WaveNet! Autoregression with long context before it was cool😅 Maybe it looks a bit silly now that we didn't use a Transformer, but we had a good reason: it would take another year for that to be invented🙃 Thanks @heiga_zen for the reminder! http

GIF

40.7K views · 501 likes · 32 reposts · 22 replies Open on X →

Sander Dieleman ✓@sedielem7 Sep

Recently, I had a great chat with my friend and former colleague @GarneloMarta, CSO at Fundamental (@LTMpredict). We talked about diffusion models (duh 🤭), music generation, typicality, blogging, @kaggle and more (also restricted Boltzmann machines 🫣) https://t.co/WFRrWI0Yhw ht

47:53

14.1K views · 191 likes · 18 reposts · 6 replies Open on X →

Sander Dieleman ✓@sedielem30 Aug

Cool to see people still remember this! Nice reminder of a fun time at the Reservoir Lab back in 2015 with @317070, @avdnoord, @lpigou, right before a bunch of us graduated and joined @GoogleDeepMind. Perhaps unsurprising that a fair few of us work on generative media today😁 14K views · 87 likes · 4 reposts · 2 replies Open on X →

Sander Dieleman ✓@sedielem30 Aug

For the curious what I'm talking about, inceptionism was this thing by @zzznah @ch402 etal: https://t.co/npndN30GoE Take an ImageNet model, fix output class and backprop to input, optimize input to maximize the output class and some smoothness auxloss. See img1. Then back at htt

34.2K views · 69 likes · 5 reposts · 3 replies Open on X →

Sander Dieleman ✓@sedielem25 Aug

@sedielem i really just want a canonical version of this to be settled on and just work out of the box, it is so obviously the right idea for a lot of important use cases. 539 views · 3 likes · 0 reposts · 1 replies Open on X →

Sander Dieleman ✓@sedielem24 Aug

@sedielem @sedielem interesting to see it making a comeback. feels like we’re on the verge of unraveling some big potentials in language AI again. 307 views · 3 likes · 0 reposts · 0 replies Open on X →

Sander Dieleman ✓@sedielem24 Aug

@sedielem I think you touched on this topic in the conversations we had at m2l summer school last september, so I was really looking forward to reading more about it from an expert like you. personally, I like that it turned into a historical account of language diffusion resear 688 views · 3 likes · 0 reposts · 1 replies Open on X →

Sander Dieleman ✓@sedielem24 Aug

New blog post: continuous diffusion for language is back! This research direction receded into the background for a while, but as of this year, it is once again a hot topic. I wrote down a historical perspective and some thoughts on the recent revival. https://t.co/gfDZgJblTu 89.4K views · 763 likes · 135 reposts · 22 replies Open on X →

Face aux comptes de taille comparable

12 posts des 90 derniers jours, à côté de la tranche de 10K–100K abonnés. diffusé largement, mais peu de ces spectateurs réagissent.

Vues médianes8 386ce compte993médiane pour 10K–100K

Portée, %11.88%ce compte3.67%médiane pour 10K–100K

Engagement, %0.70%ce compte1.61%médiane pour 10K–100K

Indicateur	Ce compte	Médiane pour 10K–100K	Rapport
Vues médianes par post	8 386	993	8.45×
Portée (vues ÷ abonnés)	11.88%	3.67%	3.24×
Taux d'engagement	0.70%	1.61%	0.44×

Autres comptes de cette tranche → Comparer avec un autre compte → Comment ces repères sont établis →

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

89.4K24 Aug

688

307

53925 Aug

34.2K30 Aug

14K

14.1K7 Sep

40.7K8 Sep

72.1K14 Sep

1.4K

1.2K

2.8K

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

1.03%24 Aug

0.58%

0.98%

0.74%25 Aug

0.23%30 Aug

0.67%

1.52%7 Sep

1.36%8 Sep

0.90%14 Sep

0.15%

0.09%

0.25%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.61%.

What the audience does

Likes57.9%2 213 in total

Reposts6.5%248 in total

Replies1.8%70 in total

Bookmarks33.8%1 292 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.

Derniers posts

Face aux comptes de taille comparable

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Comptes similaires