Yoav Gur Arieh

@GurYoav

CS PhD Student at Tel Aviv University | AI researcher at doubleAI | Researching LLM interpretability

yoav.ml Joined August 2019

188 Following 256 Followers 122 Posts

Latest posts 5

Pinned post

Yoav Gur Arieh @GurYoav May 26

Can we tell when LLMs are being unfaithful in their chains of thought?

We evaluated 8 methods claiming to do this, and found that most perform near chance!

But evaluating this requires us to have ground-truth labels for CoT faithfulness. How can we obtain these?

1:18

Yoav Gur Arieh @GurYoav Aug 25

Oh Claude...

Yoav Gur Arieh @GurYoav Jul 29

A friendly reminder to anybody writing a paper and wishing there was a way to easily get the bibtex of published versions of papers they're citing!

Yoav Gur Arieh @GurYoav May 12

Do you HATE wasting time finding the published BibTeX for a paper?

Have you always wished there would be a way to just press a button and have it copied to your clipboard?

Are you submitting to EMNLP and don't have time to spare?

Well boy have I got the chrome plugin for you!

View quoted post on X

Yoav Gur Arieh @GurYoav Jul 24

Every time

Yoav Gur Arieh @GurYoav Nov 12, 2025

When the reviewer asks for an experiment but you've already done it and it's in the appendix

View quoted post on X

Yoav Gur Arieh @GurYoav Jun 25

I spend my grant money on API credits to study the labs that fund the grants.

I call this the virtuous cycle of AI safety