Sign in

Kushin Mukherjee @ COLM

@kushinm.bsky.social
343 followers 127 following 45 posts

Postdoctoral researcher @ Stanford

PostsRepliesMedia
Reposted by Kushin Mukherjee @ COLM
Natalia Vélez @natvelali.bsky.social · 13/10/2026
Getting back into sketching this fall quarter by drawing friends old and new. :) First, notes from @lindseypowell.bsky.social's tenure talk. (Congrats Lindsey!!)
Notes from Lindsey Powell's talk, "Adventures in early social cognition". The talk describes relationships/care as a core component of intuitive psychology, and describes several experiments that show how infants and children think about social affiliation.
1475
Kushin Mukherjee @ COLM @kushinm.bsky.social · 10/10/2026
Omg congrats @veronateo.bsky.social !!!
000
Reposted by Kushin Mukherjee @ COLM
Chris Baldassano @chrisbaldassano.bsky.social · 08/10/2026
I'm recruiting PhD students this cycle! We study how the brain perceives, remembers, and predicts naturalistic experiences like movies and games, using computational tools to analyze behavioral and neural data 🧠 App deadline is 12/2/26 - see our lab pubs on our website www.dpmlab.org/pubs.html
dpmlab.org
Columbia Dynamic Perception and Memory Lab
05232
Kushin Mukherjee @ COLM @kushinm.bsky.social · 08/10/2026
They are arrived
130
Reposted by Kushin Mukherjee @ COLM
Arnav Verma @arnavverma.bsky.social · 08/10/2026
🌄 a new dawn is approaching for cognitive science 🌄 --- Lio has been one of the wisest and most thoughtful collaborators I’ve ever worked with. If you’re interested in contributing to the future of the field, look no further: web.mit.edu/zyzzyva/www/...
Text as a movie review for Arrival:
much like they are arriving
i am also arriving
@ brown university | copsy => fall 2027.
prospective grad students / postdocs / lab managers / language and movie lovers please apply! and then we can watch it all again.
web.mit.edu/zyzzyva/www/academic.html for more info
for now!
1115
Reposted by Kushin Mukherjee @ COLM
Hyowon (Hyo) Gweon @hyogweon.bsky.social · 05/10/2026
Just out in Cognition, with Aaron Chuey & @rbzsparks.bsky.social! Here we ask: What do children infer from watching others communicate? authors.elsevier.com/a/1ntbr2Hx32...
authors.elsevier.com
Please wait whilst we redirect you
All content on this site: Copyright © 2026 Elsevier B.V., its licensors, and contributors. All rights are reserved, including those for text and data mining, AI training, and similar technologies. For all open access content, the relevant licensing terms apply.
13510
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
Happening now!!
020
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
📄 Uncovering the Computational Ingredients of Human-Like Representations in Large Language Models: arxiv.org/pdf/2510.01030 📄 Evaluating Steering Techniques using Human Similarity Judgments: arxiv.org/pdf/2505.19333
arxiv.org
000
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
I'll also be doing some volunteering around the registration desks and will be around for a couple of evening socials, if folks want to talk cogsci x language models~
100
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
Next up, at poster session 4 (4:30pm – 6:30pm Wednesday), we show how most popular steering methods fail to steer model representations (and sometimes behavior!) in the same ways that simple text instructions shift people’s mental representations.
100
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
First up at poster session 1 (11am – 1pm *today*), looking at over 75 models, we investigate which ingredients (e.g., activation functions, MLP dimensionality, etc.) are associated with models whose internal representations align with those of people!
110
Kushin Mukherjee @ COLM @kushinm.bsky.social · 06/10/2026
Excited to be at @colmweb.org this week! I’m mostly going to be playing hype man for two great pieces of work on using methods from cognitive science to probe LLM representations led by the indefatigable @zstuddiford.bsky.social ! Details in thread 🧵
190
Reposted by Kushin Mukherjee @ COLM
samuel mehr @mehr.nz · 27/09/2026
another day, another bit of evidence that the scientists who willingly hand over proprietary information to AI firms may experience some consequences
7273121
Reposted by Kushin Mukherjee @ COLM
Josh de Leeuw @joshdeleeuw.bsky.social · 21/09/2026
DataPipe (pipe.jspsych.org) is ready for its post-OSF life. You can use it as before to save data from online experiments, but now the data can be sent to Zenodo, Google Drive, or a Dataverse instance. Existing OSF connections work until 11/16 I took this opportunity to do some major updates...
pipe.jspsych.org
DataPipe
13217
Reposted by Kushin Mukherjee @ COLM
Alex Zalben @azalben.bsky.social · 25/08/2026
Here's the official announcement of Parton's passing from her reps, along with the request that in lieu of flowers, the family asks for donations to be made to Dolly Parton’s Imagination Library at www.imaginationlibrary.com
NASHVILLE, TN (August 25, 2026) – Global icon and humanitarian, Dolly Parton, known to the world as a beloved singer, songwriter, philanthropist, actress, playwright, and author died peacefully today in Nashville, Tennessee. She was 80 years old.

 

Dolly is preceded in death by her beloved husband of 59 years Carl Thomas Dean (1942 – 2025), her mother Avie Lee Caroline Owens (1923 – 2003), and father, Robert Lee Parton, Sr. (1921 – 2000). 

 

A rhinestone life that shone bright enough for the world to see, Dolly will forever stand as an inspiration not only through her timeless music and prolific songwriting, but also her wit, warmth, and kindness that made us all feel like family. Dolly Parton’s legacy is one of love, compassion, and resilience. With a seven-decade career, she inspired multiple generations of artists and fans with her music and an unwavering commitment to making the world a better place. Her songs will continue to resonate with people of all ages, and her philanthropic work will have a lasting impact.

 

At Dolly’s request, a small, private service will be held for immediate family. Dolly’s fans and friends from around the world are encouraged to pay tribute to Dolly on social media by tagging @DollyParton or by leaving a message at www.dollyparton.com.         

 

In lieu of flowers the family asks for donations to be made to Dolly Parton’s Imagination Library at www.imaginationlibrary.com.
69135048673
Reposted by Kushin Mukherjee @ COLM
Jenny Saffran @jennysaffran.bsky.social · 10/08/2026
Can kids track how unfamiliar speakers use grammatical patterns to guess who is speaking? Yes they can, at least when the speakers are groups of nifty-looking aliens! Exciting new work on early sociolinguistic development by Yiran Chen and Lynna Tran in my lab. academic.oup.com/chidev/advan...
academic.oup.com
Children infer speaker groups from variation in morphosyntactic patterns
Abstract. Variation in how people use language provides powerful cues to social identity. As children join a language community, they must acquire such soc
0285
Reposted by Kushin Mukherjee @ COLM
Taylor Webb @taylorwwebb.bsky.social · 26/07/2026
I'm recruiting two postdocs (or potentially PhD students) to work on projects at the intersection of cognition, neuroscience, and AI (with a particular emphasis on mechanistic interpretability). Apply here: docs.google.com/forms/d/e/1F... and please share!
docs.google.com
37242
Reposted by Kushin Mukherjee @ COLM
Robert Hawkins @rdhawkins.bsky.social · 24/07/2026
good morning!! it's day 2 of #cogsci2026, and we're so excited to hear about new work from Seo-young Lee, Jess Mankewitz, Polina Tvilodub, Jinyi Kuang, and Karla Perez!
0228
Reposted by Kushin Mukherjee @ COLM
Arnav Verma @arnavverma.bsky.social · 22/07/2026
How do people make use of so many different modalities to communicate? Come learn more at our Unconventional Communication workshop at #CogSci2026! 🗓️ Wed, 7/22, 8:30am 🔗Full program: unconventional-communication.github.io Here's what we'll be covering 👇
unconventional-communication.github.io
Unconventional Communication (CogSci 2026)
Unconventional Communication is a workshop exploring how humans convey meaning across different modalities.
1114
Reposted by Kushin Mukherjee @ COLM
Yuka Machino @ymacci.bsky.social · 20/07/2026
Telling two people to be selfish barely changes their behavior, unless they both know they got the same message! Our #CogSci2026 paper reveals that these group expectations only matter when we all know that we share the same expectations.
1194
Reposted by Kushin Mukherjee @ COLM
Judy Fan @judithfan.bsky.social · 22/07/2026
Fri 7/24 (am): David Tomz, @kushinm.bsky.social, & co. ask about the drivers of visual variation in European Paleolithic cave art — a wonderful team spanning cognitive science & archaeology (Poster Session 2, 10am)
151
Reposted by Kushin Mukherjee @ COLM
Judy Fan @judithfan.bsky.social · 22/07/2026
Heading to #CogSci2026 in Rio this week? Overview 🧵 thread highlighting work being presented by members of the Cognitive Tools Lab (cogtoolslab.github.io) — come say hi! 🧠🛠️✨
1436
Reposted by Kushin Mukherjee @ COLM
Erik Brockbank @erikbrockbank.bsky.social · 16/07/2026
When you meet someone new, what kind of questions do you ask? It’s tempting to think that *really* getting to know people is just a matter of asking the right questions. Excited to share the next chapter in ongoing work on what makes a *deep question* with amazing collaborators...
24615
Kushin Mukherjee @ COLM @kushinm.bsky.social · 26/06/2026
Really proud to be part of this dream team! We make a strong case for end-to-end systems that can not only propose new cognitive theories but *validate* them by collecting human data! ✨ Check out project lead Ben’s thread for the highlights and a link to our preprint!
041
Reposted by Kushin Mukherjee @ COLM
Justin Yang @justintheyang.bsky.social · 13/05/2026
Thrilled to be sharing my latest work at #CogSci2026! Many of the spaces we move through, from kitchens to airports, were designed with specific uses in mind. How do people create such environments, and how do users figure out what they were designed for? 📃 osf.io/preprints/ps...
25112
Reposted by Kushin Mukherjee @ COLM
Riccardo Fusaroli @fusaroli.eurosky.social · 20/04/2026
What function did ~100k-year-old engravings from Blombos Cave & Diepkloof serve? Decoration, identity marking, proto-writing? osf.io/preprints/ps... uses transmission chains + cognitive experiments to find out & help answering one of the hardest questions in cognitive archaeology. long thread! 1/
36629
Reposted by Kushin Mukherjee @ COLM
Mike Frank @mcxfrank.bsky.social · 20/04/2026
Just wrote a new blogpost trying to summarize my thoughts on the question of how and whether to use AI for research in psychology and cognitive science: babieslearninglanguage.blogspot.com/2026/04/usin...
babieslearninglanguage.blogspot.com
Using AI to improve (not automate away) academic research
Blog about fatherhood, langauge, developmental psychology, and cognitive science.
55428
Reposted by Kushin Mukherjee @ COLM
Tobias Gerstenberg @tobigerstenberg.bsky.social · 10/03/2026
Congratulations @judithfan.bsky.social on winning the Lila R. Gleitman Prize for early-career contributions to Cognitive Science 🥳 Amazing!! cognitivesciencesociety.org/gleitman-pri...
47310
Reposted by Kushin Mukherjee @ COLM
Mike Frank @mcxfrank.bsky.social · 18/12/2025
Very excited to share the first empirical paper from LEVANTE: we describe the LEVANTE core tasks, a set of nine open source tasks for measuring learning and development in kids ages 5-12 years. osf.io/preprints/ps... 🧵
title linescreenshots from the nine tasks
29931
Kushin Mukherjee @ COLM @kushinm.bsky.social · 19/12/2025
Great post!
000
Kushin Mukherjee @ COLM @kushinm.bsky.social · 12/05/2025
Congrats!!!
020
Kushin Mukherjee @ COLM @kushinm.bsky.social · 24/04/2025
Congrats Emalie!!
010
Reposted by Kushin Mukherjee @ COLM
Erik Brockbank @erikbrockbank.bsky.social · 07/03/2025
Hello bluesky world :) excited to share a new paper on data visualization literacy 📈 🧠 w/ @judithfan.bsky.social, @arnavverma.bsky.social, Holly Huey, Hannah Lloyd, @lacepadilla.bsky.social! 📝 preprint: osf.io/preprints/ps... 💻 code: github.com/cogtoolslab/...
osf.io
OSF
4267
Kushin Mukherjee @ COLM @kushinm.bsky.social · 08/03/2025
👀📊🧠
000
Reposted by Kushin Mukherjee @ COLM
Mike Frank @mcxfrank.bsky.social · 14/02/2025
Do children understand all types of visual media in the same way? "Cross-contextual diversity in children’s early understanding of visual media" - a new review paper led by Rebecca Zhu: osf.io/preprints/ps...
Table 1 from paper: recognition and reference evidence for perception of different types of visual media by infants, toddlers, and preschoolers.
1329
Reposted by Kushin Mukherjee @ COLM
Benjamin Douglas @bendouglas.bsky.social · 13/02/2025
There are many people I’ve seen on here asking about how we go forward in the current moment We don’t have all the answers, but in the guide below my co-authors and I discuss building and sustaining academic communities dedicated to antiraicm Now at Collabra: Psychology osf.io/preprints/ps...
osf.io
OSF
175
Kushin Mukherjee @ COLM @kushinm.bsky.social · 04/02/2025
Happy CogSci evening to those who celebrate May your figure axis labels be large and visible
050
Reposted by Kushin Mukherjee @ COLM
Riccardo Fusaroli @fusaroli.eurosky.social · 19/01/2025
Why does Western paleolithic cave art strongly prefer animal side views and often abbreviations? Our new #eSymb preprint (osf.io/preprints/ps... w Pagnotta, Psujek, Mendoza Straffon and Tylén ) challenges long-held assumptions about these artistic choices based on cogsci experiments. 1/
513549
Kushin Mukherjee @ COLM @kushinm.bsky.social · 17/12/2024
Same
050
Reposted by Kushin Mukherjee @ COLM
Bria Long @brialong.bsky.social · 13/12/2024
I’m still recruiting for this position — please get in touch if you’re interested in working together!
23027
Kushin Mukherjee @ COLM @kushinm.bsky.social · 13/12/2024
Amazing title
010
Reposted by Kushin Mukherjee @ COLM
Bill Thompson @wdt.bsky.social · 13/12/2024
We have extended the deadline for this position to January 31st. Also, the start date is pretty flexible; if you're graduating next spring/summer and interested in the role, apply!
23523
Kushin Mukherjee @ COLM @kushinm.bsky.social · 11/12/2024
Wonderful! Thank you!
010
Kushin Mukherjee @ COLM @kushinm.bsky.social · 08/12/2024
Chiming in with others to say how awesome this sounds, Jen! Would also love to see the slides/materials!
120
Reposted by Kushin Mukherjee @ COLM
Daniel Martin @fizbin.bsky.social · 22/03/2024
I don't know who needs to know this, but Pro Publica has an online thing that will format a letter to your US health insurance company to demand the records behind a claim denial. (which the insurance is then legally required to provide in most cases) projects.propublica.org/claimfile/
projects.propublica.org
Find Out Why Health Insurance Denied Your Claim
You likely have the right to access records that explain why your insurer denied your claim or prior authorization request. Use ProPublica’s free tool to generate a letter requesting your claim file f...
9161794210176
Reposted by Kushin Mukherjee @ COLM
Mike Frank @mcxfrank.bsky.social · 03/12/2024
Three ManyBabies projects - big collaborative replications of infancy phenomena - wrapped up this year. The first paper came out this fall. I thought I'd take this chance to comment on what I make of the non-replication result. 🧵 bsky.app/profile/laur...
213759
Kushin Mukherjee @ COLM @kushinm.bsky.social · 05/12/2024
It’s been striking to me how much of recent air travel related issues in the northeast are due to severe air traffic controller shortages www.usnews.com/news/top-new...
usnews.com
110
Reposted by Kushin Mukherjee @ COLM
Robert Hawkins @rdhawkins.bsky.social · 01/12/2024
📄👀: cross-modal information flow in multimodal large language models neat interpretability work on how visual and linguistic information is integrated in MLLMs: "the model first transfers the more general visual features of the whole image into the representations of (linguistic) question tokens."
arxiv.org
Cross-modal Information Flow in Multimodal Large Language Models
The recent advancements in auto-regressive multimodal large language models (MLLMs) have demonstrated promising progress for vision-language tasks. While there exists a variety of studies investigatin...
0122
Kushin Mukherjee @ COLM @kushinm.bsky.social · 25/11/2024
Belongs framed on a wall much like many City Lights posters
010
Reposted by Kushin Mukherjee @ COLM
Mike Frank @mcxfrank.bsky.social · 20/11/2024
Our new paper “Is Child-Directed Speech Effective Training Data for Language Models?” was just presented at #EMNLP2024! This is work led by Steven Feng, with Noah Goodman. arxiv.org/abs/2408.03617
arxiv.org
Is Child-Directed Speech Effective Training Data for Language Models?
While high-performing language models are typically trained on hundreds of billions of words, human children become fluent language users with a much smaller amount of data. What are the features of t...
26016