Berk Ustun

@berkustun

Assistant Prof @UCSD. I work on interpretability, fairness, reliability, and personalization in ML. Previously @GoogleAI @Harvard @MIT @UCBerkeley🇨🇭🇹🇷

San Diego, CA berkustun.com Joined March 2009

981 Following 2.8K Followers 767 Posts

Latest posts 5

Pinned post

Berk Ustun @berkustun Sep 18, 2018

Denied a loan by an ML model? You should be able to change something to get approved!

In a new paper w @AlexanderSpangh & @yxxxliu, we call this concept "recourse" & we develop tools to measure it for linear classifiers.

PDF bit.ly/2xh0idm
CODE bit.ly/2xiEvls

Berk Ustun @berkustun Oct 16

4/5 of the papers in my ICLR review pile read like pure AI slop. Anyone else seeing this?

Berk Ustun @berkustun Jun 24, 2025

Explanations don't help us detect algorithmic discrimination. Even when users are trained. Even when we control their beliefs. Even under ideal conditions... 👇

Julian Skirzynski @JSkirzynski Jun 24, 2025

Right to explanation laws assume explanations help people detect algorithmic discrimination.

But is there any evidence for that?

In our latest work w/ David Danks @berkustun, we show explanations fail to help people, even under optimal conditions.

PDF shorturl.at/yaRua

View quoted post on X

Berk Ustun @berkustun Jun 21, 2024

Good news for free will!

Matthew Salganik | @msalganik.bsky.social @msalganik Jun 20, 2024

Our new paper explains why some life outcomes are hard to predict. We combined machine learning, a mathematical decomposition, and in-depth qualitative interviews to understand the origins of unpredictability: pnas.org/doi/10.1073/pn…

View quoted post on X

Berk Ustun @berkustun May 6, 2024

Our first paper on abstention as a tool for safer automation - led by @HaileyJoren! Come say hi at #ICLR24 on Thursday 🧵👇

Hailey Joren @HaileyJoren May 3, 2024

✅Easy: train a model to automate a routine task
❌Hard: ensure the model is accurate with minimal human oversight

In our latest #ICLR24, we introduce a modeling paradigm for this setting

Joint w @CharlieTMarx @berkustun

PDF:haileyjoren.github.io/files/conceptu…
ICLR:iclr.cc/virtual/2024/p…

View quoted post on X