Emma Pierson

Our paper, "What's in My Human Feedback", received an oral presentation at ICLR!

Our method automatically+interpretably identifies preferences in human feedback data; we use this to improve personalization + safety.

Reach out if you have data/use cases to apply this to!

arxiv.org/pdf/2510.26202

New research is offering new insight on how Americans move — all the way to the neighborhood level.

A new dataset, MIGRATE, maps annual moves with 4,600‑times more detail than standard public data, revealing patterns hidden in county‑level reporting:

https://bit.ly/49XSD6w

Paper: nature.com/articles/s4146
7-025-68019-2

Cornell Chronicle article: news.cornell.edu/stories/2026...

Data access: migrate.tech.cornell.edu

Joint work led by the geospatial wizard @gsagostini.bsky.social and with great coauthors Rachel Young, Maria Fitzpatrick, and @nkgarg.bsky.social!

Now out in Nature Communications - we have released a migration dataset that is

- 4000x more granular than existing public data
- highly correlated with Census data
- being used by >100 academic, govt, and non-profit teams all over the world

See @gsagostini.bsky.social's thread!

This work is led by the wonderful James Diao, with a great team of coauthors: @rajmovva.bsky.social, Lingwei Cheng, @kkado.bsky.social, Aashna Shah, @neil-r-powe.bsky.social, Kadija Ferryman, and Raj Manrai!

Public Opinion on Use of Race in Clinical Algorithms This survey study assesses public opinion and preferences among US adults regarding the use of race in clinical algorithms.

See the paper - jamanetwork.com/journals/jam... - for more details, including sensitivity analyses, replication of prior gold-standard surveys, etc!

Full survey questions, data, and code: github.com/epierson9/ra...

Finding #4: Respondents were 4x more likely to be uncomfortable if clinicians used race without asking...yet <10% reported ever being told their race was used.

This suggests the way we communicate about the use of race may not foster trust, and raises concerns in light of calls for transparency.

Finding #3: Respondents were more comfortable with the use of race than with widely-proposed alternatives like zipcode or income.

Said one respondent: "What does my paycheck have to do with a genetic mutation?"

Don't assume switching to these factors will automatically improve trust.

Finding #2: However, a substantial minority of respondents were not comfortable with use of race, and Black + Hispanic respondents were less comfortable than white and Asian respondents.

This raises complex ethical and algorithmic questions about how to weigh these clashing preferences.

Finding #1: Most respondents were comfortable with the use of race in at least some circumstances.

This highlights a gap between calls to eliminate uses of race in medicine and public opinion.

We have a new paper in JAMA Internal Medicine!

Patient race is widely used in medical algorithms...but it's unclear how patients feel about this.

We conduct the first nationally-representative YouGov survey to find out, producing four findings with practical clinical implications. 1/

AI has a bias problem. Can we build something smarter? - Berkeley News UC Berkeley computer scientist Emma Pierson believes we can use AI to improve our healthcare and criminal justice systems — but only if we design these algorithms with an eye toward equality.

Thanks to Kara Manke at Berkeley News for this profile of our lab's recent work on fairer decision-making in healthcare and policing! news.berkeley.edu/2026/01/20/a...

Thanks - super-interesting, and actually very relevant to some other work we're doing as well. Will pass along!

We're excited about applications of our test to other datasets that have 1) perceptions of race, gender, etc and 2) multiple observations of the same person.

This work is led by the wonderful Nora Gera, in a great start to her PhD!

Full paper: www.science.org/doi/epdf/10....

See the paper for many robustness checks and discussion of nuances! Our finding persists when using alternate outcomes, statistical models, subsets of the data, and controls satisfying the criteria above.

5/

A benefit of our test is that it doesn't require us to control for all factors legitimately influencing searches. We only have to control for things that influence both searches and perceived race, vary for the same person across stops, and don't themselves suggest bias.

4/

9% of drivers stopped multiple times have inconsistently perceived race across different stops - most perceived as both white + Hispanic.

When perceived as Hispanic, the same driver is likelier to be searched/arrested. This gap is substantial (24% of overall search rate).

3/

Tests for racial bias often compare how two people of different races are treated.

But two people typically differ in many ways besides race.

So instead of comparing two different people, we study the *same person over time*, as perceptions of their race change.

2/

We have a new paper in Science Advances proposing a simple test for bias:

Is the same person treated differently when their race is perceived differently?

Specifically, we study: is the same driver likelier to be searched by police when they are perceived as Hispanic rather than white?

1/

New #NeurIPS2025 paper: how should we evaluate machine learning models without a large, labeled dataset? We introduce Semi-Supervised Model Evaluation (SSME), which uses labeled and unlabeled data to estimate performance! We find SSME is far more accurate than standard methods.

selfishly i wish we could keep divya in our lab forever but i guess it would be a disservice to the rest of the world 😅 she’s been such a wonderful mentor to me—i’ve learned a lot from how thoughtful, creative, and knowledgeable she is about everything. she’s also super funny and amazing at baking 🤭

Meeting Divya 5 years ago was one of the biggest strokes of luck in my faculty career - she is a brilliant scientist who has been foundational to so many of our lab's projects, and any institution would be lucky to hire her.

Postdoctoral Employee - Artificial Intelligence - Electrical Engineering and Computer Sciences Department University of California, Berkeley is hiring. Apply now!

Apply here - aprecruit.berkeley.edu/JPF05028 by 11/15, but review of applications is ongoing so sooner is better! (Application deadline currently says 9/15 but will be extended).

Sparse Autoencoders for Hypothesis Generation We describe HypotheSAEs, a general method to hypothesize interpretable relationships between text data (e.g., headlines) and a target variable (e.g., clicks). HypotheSAEs has three steps: (1) train a ...

Broad project areas include:

1) language modelling methods for scientific discovery (building on our recent work - arxiv.org/abs/2502.04382)

2) using language models to support equity (ai.nejm.org/doi/full/10....)

both in collaboration with health+social scientists.

2/3

🚨 New postdoc position in our lab at Berkeley EECS! 🚨

(please reshare)

We seek applicants with experience in language modeling who are excited about high-impact applications in the health and social sciences!

More info in thread

1/3

📢New POSITION PAPER: Use Sparse Autoencoders to Discover Unknown Concepts, Not to Act on Known Concepts

Despite recent results, SAEs aren't dead! They can still be useful to mech interp, and also much more broadly: across FAccT, computational social science, and ML4H. 🧵

SF fog coming up to swallow us in time lapse.

Honored to win a #CHIL2025 best paper award for our work modeling inequality in disease progression, led by @ericachiang.bsky.social!

To the NIH: health inequality remains a vital topic to support the health of all Americans. As we prove, failing to account for it biases estimates for everyone.

For folks at @facct.bsky.social, our very own @cornellbowers.bsky.social student @emmharv.bsky.social will present the Best-Paper-Award-winning work she led on Wednesday at 10:45 AM in the "Audit and Evaluation Approaches" session!

In the meantime, 🧵 below and 🔗 here: arxiv.org/abs/2506.04419 !

assassinations, handcuffing a senator at press conference, marines detaining a civilian, and a military parade for the president’s birthday. rough week for democracy.