Francis Bach

@BachFrancis · Paris, France · joined 01 May 2013

Researcher in machine learning

17 829Followers

95Following

361Posts total

56.8KViews on collected posts

Derniers posts

Francis Bach@BachFrancis20 Jul

L’Inria et le CNRS toujours au top, cette fois-ci avec nos amis de Grenoble. 170km, 5200m de dénivelé positif pour une Etape du Tour autour de l’Alpe d’Huez. Bravo à @julienmairal, @olivier_cappe et @JulyanArbel. https://t.co/0sBfbsymRF

4.8K views · 49 likes · 3 reposts · 1 replies Open on X →

Francis Bach@BachFrancis24 Mar

Great! Time to sharpen our pencils for some great submissions to present in Paris. 5.4K views · 44 likes · 0 reposts · 0 replies Open on X →

Francis Bach@BachFrancis5 Mar

Looking for alternatives to quadratic functions for closed-form analysis in optimization? This post explores matrix Riccati dynamics and their applications to neural networks. https://t.co/nEJdNGnNEW https://t.co/nDsjJ8OBuP

GIF

9.7K views · 163 likes · 23 reposts · 0 replies Open on X →

Francis Bach@BachFrancis27 Sep 2025

Together with the great Frederik Kunstner (https://t.co/PP2JY3QqK4) 3.6K views · 21 likes · 2 reposts · 0 replies Open on X →

Francis Bach@BachFrancis27 Sep 2025

Not all scaling laws are nice power laws. This month’s blog post: Zipf’s law in next-token prediction and why Adam (ok, sign descent) scales better to large vocab sizes than gradient descent: https://t.co/uoy5GPrZek https://t.co/gQ9DqwtBsF

GIF

29.5K views · 459 likes · 49 reposts · 4 replies Open on X →

Francis Bach@BachFrancis21 Jul 2025

Jamais deux sans trois. Le département d’informatique de l’ENS toujours en forme. 131 km, 4500m de dénivelé, cette fois-ci avec du Beaufort aux ravitaillements! Bravo @olivier_cappe! #LEtapeduTour https://t.co/H0rRx7G8V2

3.8K views · 61 likes · 0 reposts · 3 replies Open on X →

Face aux comptes de taille comparable

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

Vues médianes4 826ce compte940médiane pour 10K–100K

Portée, %27.07%ce compte3.45%médiane pour 10K–100K

Engagement, %1.12%ce compte1.55%médiane pour 10K–100K

Indicateur	Ce compte	Médiane pour 10K–100K	Rapport
Vues médianes par post	4 826	940	5.13×
Portée (vues ÷ abonnés)	27.07%	3.45%	7.85×
Taux d'engagement	1.12%	1.55%	0.72×

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

3.8K21 Jul

29.5K27 Sep

3.6K

9.7K5 Mar

5.4K24 Mar

4.8K20 Jul

Last 6 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

1.70%21 Jul

1.74%27 Sep

0.64%

1.92%5 Mar

0.82%24 Mar

1.12%20 Jul

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.55%.

What the audience does

Likes63.0%797 in total

Reposts6.1%77 in total

Replies0.6%8 in total

Quotes0.2%3 in total

Bookmarks30.1%381 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.