Xuechunzi Bai @baixuechunzi Twitter profile | Pikagi

Pikagi

Xuechunzi Bai

@baixuechunzi

864

Followers

633

Following

11

Media

162

Statuses

Psychologist studying social cognition, stereotypes, individual & structural, computational principles, in human and AI. Assistant Prof @UChicago

Hyde Park, Chicago

https://t.co/Cwb6bU9Cc1

Joined May 2012

Don't wanna be here? Send us removal request.

Pinned Tweet

@baixuechunzi

Xuechunzi Bai

2 months

🚨 New Preprint 🚨 OpenAI says that the new GPT-4o model is carefully designed and evaluated to be safe and unbiased. But implicit biases often run deep and are harder to detect and remove than surface-level biases. In our preprint, we ask, are GPT-4o and other LLMs really

Tweet media one

14

103

516

Last Seen Profiles

@Crippledbtc

@alhaziyn

@tsubokuraa65043

@AraOshagan

@elinadaun

@NamsCompendium

@kiuridaisuki_ch

@Steven_Weinberg

@BCS_dnd_club

@burrowsvech

@scottwbarrow

@7EENE

@MechaWaffle3

@HallidayMe24771

@wannsato

@JamesHump8822

@siva07270_

@stwmaniax

@EvilScaryFruits

@ALyahia22Zakray

@xRealCulture

@norhan_kelany

@H43522819

@r3dpanama

@nanevl

@vfKxT8rn5lnBeJM

@PaperCrownRw

@galicia_sa45256

@teamissue

@nicolemckenz

@m7amadm1984

@demishassabis

@kanavnee

@swryt84639

@Himna_

@slmnlsydy1

@baixuechunzi

Xuechunzi Bai

3 years

Psychological Science in press w/ Susan Fiske, Tom Griffiths An undesirable outcome can come from what seems like a rational process. 1/9

Globally Inaccurate Stereotypes Can Result from Locally Adaptive Exploration

Inaccurate stereotypes -- perceived differences among groups that do not actually differ -- are prevalent and consequential. Past research explains stereotypes as emerging from a range of factors...

2

16

57

@baixuechunzi

Xuechunzi Bai

9 months

Grad students 👏 Please apply to become a VIP at Princeton Summer 2024 🤩 Visiting Internship for PhD students (VIPS). More info below.

Visiting Internship for Ph.D. Students (VIPS) Program

The VIPS program is a new initiative of the DEI committee of the Department of Psychology at Princeton University. VIPS will support 2-4 Ph.D. students who are full-time students at other universit...

psych.princeton.edu

0

23

54

@baixuechunzi

Xuechunzi Bai

2 years

Loved this initiative! but 😂

Tweet media one

@StacyTShaw

Dr. Stacy T. Shaw

2 years

Need help pronouncing Chinese student names in your class? The amazing @xiwen_lu has created a tool where you can look up any name and hear the most common pronunciation!

Tweet media one

35

1K

6K

2

0

29

@baixuechunzi

Xuechunzi Bai

2 years

@Ivuoma Thanks Ivy! I am Bai, a rising 5th year at Princeton psych, policy, stats and machine learning. I study social stereotypes. My job talk features a new theory on the origin of immigrant stereotypes, and offers implications for structural diversity.

www.xuechunzibai.com

0

3

29

@baixuechunzi

Xuechunzi Bai

2 years

people evaluate social groups with **many** dimensions, going beyond mere valence (good v bad), the big two (warmth/competence), or three (ideology). natural language illustrates!

@NicolasGandalf

Gandalf Nicolas

@NicolasGandalf

2 years

New paper! @baixuechunzi , Susan Fiske, and I introduce the Spontaneous Stereotype Content Model (SSCM). We use various text analysis methods in combination with free responses to propose a comprehensive taxonomy of spontaneous stereotype content.

4

33

121

0

1

25

@baixuechunzi

Xuechunzi Bai

2 months

2/ We introduce two measures of bias, inspired by psychological methods to tackle similar issues in humans: 💡LLM Implicit Bias: A prompt-based approach to uncover hidden biases. 💡LLM Decision Bias: Detect subtle discrimination in decision-making tasks.

Tweet media one

2

5

23

@baixuechunzi

Xuechunzi Bai

9 months

👏🤩

@LydiaFEmery

Lydia Emery

9 months

The University of Chicago social psych area is hiring at the TT assistant professor level! Please apply and help us spread the word as we continue to build up our social psychology area. Job ad here; review begins 10/23:

Tweet media one

3

73

145

0

0

17

@baixuechunzi

Xuechunzi Bai

2 months

6/ While significant progress has been made in reducing stereotype biases in LLMs, there is still much to be learned from the origin of these biases: humans. Stay tuned for more updates from us with @ang3linawang , @sucholutsky , @cocosci_lab ! (old) preprint:

1

2

15

@baixuechunzi

Xuechunzi Bai

2 months

4/ Let's spotlight the racism test. Here, GPT-4 assigns all 8 positive words to white and all 8 negative words to black. Although human participants also tend to associate the concept of black with negativity, it is not to the same levels of GPT-4’s confidence (no uncertainty)

6

3

14

@baixuechunzi

Xuechunzi Bai

2 months

5/ This pattern isn't just a one-off error but a consistent trend in the 8 tested models, including GPT-4o which claimed to have “... undergone extensive external red teaming with 70+ external experts in domains such as social psychology, bias and fairness, ...”.

Tweet media one

1

0

13

@baixuechunzi

Xuechunzi Bai

1 year

Maybe also The World Map according to the data *psychologist* sees

@vdignum

Virginia Dignum is mostly commenting on #LinkedIn

1 year

The world according to the data #AI sees. 50% of datasets are connected to 12 institutions. Analysis of 4384 datasets and 60647 papers. Read the report: By @bernardkoch @emilydenton77 @alexhanna

Tweet media one

37

571

1K

0

0

13

@baixuechunzi

Xuechunzi Bai

2 months

1/ Large Language Models (LLMs) like GPT are great at passing explicit bias tests, but they might still have implicit biases, similar to humans who espouse egalitarian beliefs yet exhibit subtle biases. How can we systematically measure these implicit biases?🤔

1

1

13

@baixuechunzi

Xuechunzi Bai

2 years

Very excited to host the #SPSP2022 symposium on computational social cognition with @fierycushman @tatianalau @minjaejk and Susan Fiske! Friday at 3:30 PT online. Enjoy the intro & Hope to see you there!

0

6

12

@baixuechunzi

Xuechunzi Bai

2 months

3/ Our studies with 8 value-aligned LLMs across race, gender, religion, and health reveal pervasive human-like stereotypes in approx. 20 categories, including criminality and race, science and gender, valence and religion, disability, age, etc.

Tweet media one

1

3

10

@baixuechunzi

Xuechunzi Bai

3 years

Pragmatically, our paradigm hypothesizes desegregated diversity at the level of the environment, and humble exploration at the level of individual, matter. See also paper with Susan Fiske and Miguel Ramos on diversity and stereotype dispersion. 8/9

Tweet media one

1

0

7

@baixuechunzi

Xuechunzi Bai

2 years

@olsonista my name prob is an outlier cuz it has more characters than a typical Chinese name. maybe the algo detected my family’s intention to squeeze my Korean, Japanese, and Chinese ancestry into one name! lol

0

0

6

@baixuechunzi

Xuechunzi Bai

3 years

To share a paper which also provides great counter-argument; we need a social constructivism viewpoint:

Tweet card media

'Thinking ill of others without sufficient warrant?' Transcending the accuracy-inaccuracy dualism...

Research on prejudice seeks to understand and transform inaccurate beliefs about others. Indeed, historically such research has offered a cautionary tale of the biased nature of human cognition....

pubmed.ncbi.nlm.nih.gov

@spiantado

steven t. piantadosi

3 years

Did you know that there are psychologists who study "stereotype accuracy"? I've always wondered what the hell, so I've been reading it recently. For the record, it’s exactly as bad as it sounds. Here’s a thread.

44

622

4K

0

0

6

@baixuechunzi

Xuechunzi Bai

3 years

The minimal process is local, adaptive exploration: Formally, multi-armed bandit models show that the mere act of choosing among groups with the goal of maximizing long-term benefits -- when all groups are equally, highly rewarding -- is enough. 4/9

Tweet media one

Tweet media two

1

0

4

@baixuechunzi

Xuechunzi Bai

3 years

Inaccurate stereotypes about social groups are widespread and consequential, but their origin is puzzling. People often think social groups differ from each other, even absent group-level differences. We propose a minimal, functional paradigm that suffices to produce bias. 3/9

Tweet media one

1

0

4

@baixuechunzi

Xuechunzi Bai

3 years

Inaccurate stereotypes don’t require group-serving motivations, cognitive limitations, or information deficits? Locally adaptive exploration can produce globally inaccurate judgments. 2/9

1

0

4

@baixuechunzi

Xuechunzi Bai

3 years

This work benefited tremendously from great discussions with members at the fiske lab, at cocosci lab, at tlab, at social psychology brown bag, at bias in machine and in human datax workshop, as well as editors and reviewers at PsychSci. A big thank-you! 9/9

0

0

4

@baixuechunzi

Xuechunzi Bai

9 months

@LydiaFEmery 🥳🎉🎊

0

0

3

@baixuechunzi

Xuechunzi Bai

3 years

@BareketOrly Thank you Orly!!

0

0

2

@baixuechunzi

Xuechunzi Bai

3 years

Theoretically: Perhaps one origin of stereotypes is much simpler than we have thought; even minimal assumptions can recreate it. However, this simplicity is also troubling: 6/9

1

0

2

@baixuechunzi

Xuechunzi Bai

3 years

Empirically, this phenomenon is reproduced in two large online experiments (N = 2404). Please feel free to play around with the code, data, experiments, and reach out! 5/9

Tweet media one

Tweet media two

1

0

2

@baixuechunzi

Xuechunzi Bai

2 months

@DanHendrycks @sucholutsky @GaryMarcus @MelMitchell1 @mmitchell_ai @emilymbender @savvyRL @IasonGabriel @manish_raghavan @aylin_cim Stay tuned & Releasing soon!

0

0

1

@baixuechunzi

Xuechunzi Bai

3 years

If stereotypes can result from each person pursuing their own self-interest, we may need to work harder to create environments where problematic stereotypes do not develop. 7/9

1

0

1

@baixuechunzi

Xuechunzi Bai

3 years

@fasbrock @SocietiesHybrid @TUChemnitz Thanks Frank for hosting this fascinating talk series! Our great pleasure to be able to join the conversation and to keep learning!

1

0

1

@baixuechunzi

Xuechunzi Bai

2 years

@paulrconnor Interesting! The analysis should be fine but getting high quality data might be tricky.

1

0

1