Ghassen Jerfel

Research @ Anthropic. 🇹🇳 ✈️ 🇺🇸 Former @Waymo @GoogleBrain, @DukeU, @UCBerkeley, @Princeton.

[Andrej Karpathy @karpathy]

This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads.

Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.

[Jeff Dean @JeffDean]

Waymo's system, fueled by careful collection of a large volume of fully autonomous data, is the most advanced, large-scale application of embodied AI today. Very proud to see this level of engineering rigor tackling safe autonomous driving making the roads safer for everyone (and it has been nice to see various Google research collaborations with Waymo be a part of these advances!).

The insights here are foundational for how we design and safely scale all complex AI systems.

Read more at:
https://t.co/TFk9yv6rce

[Ghassen Jerfel @Ghassen_ML]

@deepcohen @jasondeanlee @SebastienBubeck @tomgoldsteincs @zicokolter @atalwalkar Congratulations Jeremy!!!

[Ghassen Jerfel @Ghassen_ML]

@XYHan_ @ChicagoBooth Congrats !!!

[Daylen Yang @daylenyang]

40 minutes of fully autonomous driving in a @Waymo 🤖🚗 with no one up front! Most people didn’t notice at all, but those that looked in did a huge double take 🙂

[Ghassen Jerfel @Ghassen_ML]

@theplankt “Extravagant” seals the deal tbh

[Dustin Tran @dustinvtran]

In our work "Plex", we propose a framework for reliability in AI. We also introduce new models (ViT-Plex & T5-Plex) for reliable decision-making across a broad array of scenarios.

Blog: https://t.co/bCkWQ8C2qM
Paper: https://t.co/OiYWv28uwa
Code: https://t.co/xz3FAzBxdq

[Ghassen Jerfel @Ghassen_ML]

@bidumit @BeEngelhardt @Columbia @Cancer_dynamics Congratulations Bianca!!

[Ghassen Jerfel @Ghassen_ML]

@AaronSchein @UChicago @DSI_UChicago Congrats Aaron!!!

[Ghassen Jerfel @Ghassen_ML]

@roydanroy Used to be AISTATS.

[Jasper @latentjasper]

Excited about this paper from my colleagues (https://t.co/Zildpnp27O) which builds a formal understanding of the behavior of overparameterized models under covariate shift. Amazingly, the theory lines up surprisingly well with what we observe with large deep nets.

[Ghassen Jerfel @Ghassen_ML]

@BeEngelhardt @Princeton Congratulations!!!

[Ghassen Jerfel @Ghassen_ML]

@Hansenq @Jabra_US @wirecutter Get the memory foam tips. That was the only way I could keep my 65t in place especially when running.

[Ghassen Jerfel @Ghassen_ML]

@kat_heller Merry Christmas !

[Behnam Neyshabur @bneyshabur]

If you believe you are at a disadvantage in ML community (e.g. because of your race, gender, nationality, background or other circumstances) and need guidance & help, I'd love to meet you! Just pick a time from here:

https://t.co/GHTL4oivqu

Please RETWEET to spread the word!

[Ghassen Jerfel @Ghassen_ML]

@theplankt Sodastream’s fizziness can be weak and their CO2 tank refills are a bit pricy. A DIY machine can be quite easy to build and you can use cheap paintball CO2 tanks. You can also easily control fuzziness with a knob ;)

[Ghassen Jerfel @Ghassen_ML]

@theplankt you can also build one yourself (to ensure max fizzy)

[Ghassen Jerfel @Ghassen_ML]

@leezabuh Oof this pic reminds me of the train wreck that has been my cardiovascular health since the start of lockdowns.

[Ghassen Jerfel @Ghassen_ML]

@theplankt I just find myself making plans and summaries with alternating pens but that just makes me write rather than act on my plan ...