Transparency for foundation models is an outstanding challenge.

To make progress the White House and G7 have recommended that foundation model developers prepare *transparency reports*.

We recently put out a paper that articulates what this should mean and its policy impact🧵

Me too! I am lucky to have read it early and I heartily endorse it.

i'm so hyped about this book. just pre-ordered it. you should too

Growing evidence has revealed deep flaws in how machine learning is used in science, a problem that spans dozens of fields.

New guidelines from an interdisciplinary team, including Arvind Narayanan, Sayash Kapoor and Emily Cantrell, tackle this problem.

bit.ly/49YBnuA

'AI now beats humans at basic tasks': Really?

open.substack.com/pub/aiguide/p/…

Very interesting paper on overreliance in LLMs, led by Sunnie S. Y. Kim.

The results on overreliance are very interesting, but equally fascinating is the evaluation design: they random assign users to different LLM behaviors + check against a baseline with internet access.

There is a lot of interest in estimating LLMs' uncertainty, but should LLMs express uncertainty to end users? If so, when and how?

In our #FAccT2024 paper, we explore how users perceive and act upon LLMs’ natural language uncertainty expressions.

arxiv.org/abs/2405.00623

1/6

Very proud to share that our paper introducing the REFORMS checklist is now published in Science Advances! Within this paper, we propose a checklist of 32 questions across 8 different steps of an ML pipeline that should help avoid common mistakes.

science.org/doi/10.1126/sc…

Great to have been involved in this initiative led by Sayash Kapoor and Arvind Narayanan to (hopefully!) improve the use of machine learning in science. Further thoughts in my Substack post: fetchdecodeexecute.substack.com/p/reforms-a-gu…

REFORMS focuses on applying ML in the sciences; good toh highlight some folks within ML who have worked on reproducibility for a long time:

Joelle Pineau - cs.mcgill.ca/~jpineau/Repro…

Percy Liang - worksheets.codalab.org

Jesse Dodge - jessedodge.github.io/NLP_Reproducib…

REFORMS is an exceptional work by an ensemble cast spanning institutions and disciplines! Check it out!

The approach also directly inspired our work on open foundation models, where we worked towards consensus across folks from different institutions:
crfm.stanford.edu/open-fms/

Check out our new paper! Thanks to Sayash Kapoor for leading this amazing project.

Lots of practical advice to help researchers doing ML-based science avoid unintentional irreproducibility and overgeneralization in this new paper led by Sayash Kapoor

I am really excited to be part of this project led by Sayash Kapoor and Arvind Narayanan to help improve practices in machine-learning based science. science.org/doi/10.1126/sc…