I've updated the course notes for the grad stats () class I help teach here at Stanford.
You can find them compiled into a 📓 here:
The lecture slides that go with the notes are here:
#RStats
Excited to gear up for another round of teaching my grad stats class at Stanford!
📖 You can find the course notes compiled into an online book here:
🛝 And the lectures slides that go with the notes here:
🚨 New preprint 🚨
In "Beyond the here and now: Counterfactual simulation in causal cognition", I discuss what role counterfactual simulation plays for how people judge causation and assign responsibility.
📰
How do people make causal judgments? 🤔👉💨🎱
"A counterfactual simulation model of causal judgment" is finally out as a preprint (~ 9 years in the making!) with Noah Goodman, David Lagnado, and Joshua Tenenbaum.
📰
📎
1/7
I'm so honored to receive this prize from the
@socphilpsych
! I'm grateful to my mentors and all my colleagues 🙏 Special thanks to
@Ericmandelbaum
and
@phillipsjs
for pulling off an excellent zoom surprise 🤓
Can't wait to see many of you at the 50th SPP conference in Purdue ♥️
I'm delighted to announce the 2024 winner of the Stanton prize is…
@tobigerstenberg
! The
@socphilpsych
Stanton Prize is awarded to a young scholar who has made significant contributions to interdisciplinary research & been active in the SPP. Please join me in congratulating Tobi!
The talk recordings from the "Causality in Minds and Machines" workshop are live now.
Thanks again to all the speakers and to
@socmathpsych
for sponsoring the workshop!
🚨 New preprint 🚨
In "Beyond the here and now: Counterfactual simulation in causal cognition", I discuss what role counterfactual simulation plays for how people judge causation and assign responsibility.
📰
How do people make causal judgments? I argue that counterfactuals are key & hypotheticals don't suffice.
If you're like counterfactuals, hypotheticals, hipnoschweticals, what's the difference? 🙄This preprint is for you!
📰
📺
🧵👇
Now out in JEP:G. "Inference from Explanation" by
@Lara_Kirfel
w Thomas Icard & myself.
📰
📜
Also, I'm excited to announce that Lara will be joining the CiCL as a postdoc 🥳🎈🌴 I'm looking forward to many more projects together!
How do people make causal judgments? 🤔👉💨🎱
"A counterfactual simulation model of causal judgment" is finally out as a preprint (~ 9 years in the making!) with Noah Goodman, David Lagnado, and Joshua Tenenbaum.
📰
📎
1/7
I'm excited to jointly teach "What makes a good explanation? Psychological and philosophical perspectives" with Thomas Icard starting next week.
Check out the syllabus here
What papers are we missing? 🤔
Hi causality fans! I'm pleased to announce our
@socmathpsych
workshop on causality in minds 🤔 and machines 🤖 in San Francisco on November 16th.
We have an amazing lineup of speakers, and we invite poster presentations (workshop is free)!
Infos here:
Make the beginning of your zoom lecture less awkward by
🎶 playing music from a collaborative spotify playlist that students can contribute songs to
💬 adding a prompt on something to chat about
🖼️ showing something fun/nice to look at
🤓📣 Now out in Psychological Review (
@APA_Journals
): "A counterfactual simulation model of causal judgments for physical events" with Noah Goodman,
@david_lagnado
& Joshua Tenenbaum.
📰:
🗞️pre-print:
📎:
Hot off the press: Our paper in Psychological Science shows how people spontaneously engage in counterfactual simulation when making causal judgments.
@PsyArXiv
preprint and paper
Excited to introduce the 'Counterfactual Simulation Model of Causal Language' 🎱➡️💬 (led by Ari Beller, ) which captures how people's causal representations of what happened translate into language.
📰
📎 🧵
I met one of my intellectual heroes for lunch today!
Phil Johnson-Laird () develops mental model theory -- an influential framework that explains how people reason.
I expected that Phil would be smart & wise. Turns out he's incredibly funny, too!
The recording of my HAI seminar talk "Counterfactual simulation in human cognition " is now available online here:
Thanks
@StanfordHAI
for having me, and thanks everyone for joining in!
👇: 🗼🤓📺
New preprint! "Mental Jenga: A counterfactual simulation model of physical support" by
@itsaliang
with
@realkevinsmith
& Josh Tenenbaum.
Most work on causal judgments focuses on dynamic events. Support is different: it's causal but nothing happens!
🧵
New preprint! "Mental Jenga: A counterfactual simulation model of physical support" by
@itsaliang
with
@realkevinsmith
& Josh Tenenbaum.
Most work on causal judgments focuses on dynamic events. Support is different: it's causal but nothing happens!
🧵
What did I see?
A fantastic keynote by
#SPP2022
Stanton prize winner
@chazfirestone
on how perception is much more than seeing 'what' is 'where'.
People also see 'how'! How things fit, relate, unfold over time, ...
Congrats Chaz!!
Sarah Wu's poster on how people assign responsibility to helping and hindering agents: Together with Shruti Sridhar she develops a computational model that combines counterfactual simulation & intention inference to predict responsibility.
#CogSci2023
New preprint ⏰ In "Making a difference: Criticality in groups"
@RoiZultan
,
@david_lagnado
and I study people's judgments of how critical individual group members are for joint outcomes.
📰
📎
🧵1/7
📣New preprint! In "If not me, then who? Responsibility and replacement" Sarah Wu () shows that responsibility judgments in groups relate to how easily a contribution could have been replaced.
📄:
💻:
1/7
Responsibility judgments are driven by causal attributions and mental state inferences. Sarah Wu () & Shruti Sridhar develop a computational model demonstrating how these components work and connect.
📰
📎
Powerful talk by
@emilymbender
at
#CogSci2022
on "Resisting dehumanization in the Age of AI" with concrete suggestions of what cognitive scientists can do.
New paper out in
#JEPG
with Thomas Icard: "Expectations affect physical causation judgments". First 📰with
@Stanford
affiliation🍾🤓
Paper:
Pre-print:
Materials:
Thread 1/7
Sarah Wu () just gave an amazing first-year-project talk today on the role of counterfactual reasoning in responsibility judgments!
Sarah studies what counterfactuals come to mind when we evaluate others' actions. Catch her at SPP & CogSci!
🧑🦰💭🐖🎈
📣New preprint! In "If not me, then who? Responsibility and replacement" Sarah Wu () shows that responsibility judgments in groups relate to how easily a contribution could have been replaced.
📄:
💻:
1/7
Very proud of my student Sarah Wu () who gave an awesome talk in our joint cog & neuro seminar. She presented a computational model that predicts responsibility judgments in social scenarios grounded in counterfactual simulations and intention inferences.
What do we learn from causal explanations?
In a new pre-print with Thomas Icard and
@tobigerstenberg
, we find that causal explanations disclose much more than what is explicitly stated - “Inference from explanation”:
A thread 🧵⬇️
#norms
#causality
"Psychology is about understanding human behavior; Law is about regulating human behavior; You'd think they have a lot to say to each other."
In an equally insightful & funny keynote at
#CogSci2023
@BobbiePoPS
points out the many ways in which 👩⚖️ & 🤓 can learn from each other.
Very excited to hold
@nickchk
's "The effect" in my hand (depicted for scale)! Just in time to update my grad stats class here at Stanford:
free online here 👉
Very excited to start working on this with
@jiajunwu_cs
,
@chelseabfinn
, Noah Goodman, Thomas Icard, and Rob MacCoun. Thanks for the support from
@StanfordHAI
!!
Stay tuned for updates as we develop MARPLE 🕵️♀️
Scholars will develop MARPLE, a computational framework that combines evidence from vision, audio, and language to develop human-understandable explanations. This could improve home assistants, support legal fact-finders, and even help us better understand human inference. (6/7)
Thought-provoking 🤯 and funny 😂 talk by
@YejinChoinka
on modeling common sense, arguing for the importance of language and causality, and questioning GTP3's moral reasoning capabilities 🤔 at
#NeuroHAI
conference.
I'm so looking forward to reading
@david_lagnado
's upcoming book "Explaining the evidence" 🕵️
Dave is the most wonderful scientist, writer, advisor, and story-teller. The book will be a treat!
👇 Collage for a thank-you mug that sits in Dave's office :)
I'm excited to give a talk tomorrow at 10am PT as part of the
@StanfordHAI
seminar series. I'll be talking about the role of counterfactual simulation in human cognition.
Join me here and ask questions here
Check out Sarah Wu's () poster "That was close! A counterfactual simulation model of causal judgments about decisions" (P2-19-2241) in tomorrow's session from 1pm to 3pm @
#CogSci2022
Thanks
@mjskay
for a fabulous guest lecture in our graduate stats class here
@Stanford
(), showing us how to effectively visualize uncertainty. And thanks to the tidybayes package () we can all do it, too! 📈👍
#rstats
Why did I do that? 🤔 Today in lab meeting we discussed "Rationalization is rational" (). Thanks
@fierycushman
for a thought-provoking and beautifully written paper!
How many parameters do YOU need to explain memory?
@timothyfbrady
needs1⃣ At
@StanfordPsych
, Tim argued for continuous, population-based memory representations, and against the idea that we can only hold N discrete items in mind.
much more here 👉📰:
Congrats to our wonderful cohort of
@StanfordPsych
honors students 🎈🌴
I hope you'll all have a wonderful summer and I wish you all the best for your next 👟👟
Are you the next Sherlock 🕵️♂️or Miss Marple 🕵️♀️? Check out our work that looks at how people combine evidence from vision 👀 and sound 👂to figure out what happened and who did it.
Brought to you today as poster P1-LL-148 at
#CogSci2024
New preprint ⏰ In "Making a difference: Criticality in groups"
@RoiZultan
,
@david_lagnado
and I study people's judgments of how critical individual group members are for joint outcomes.
📰
📎
🧵1/7
How do people figure out what happened in the past?
Ari Beller, Yingchen Xu,
@scott_linderman
and I use 👀-tracking to show that people sometimes do so through mental simulation.
📰
📎
🧵 1/8
#CogSci2022
#SPP2022
Thanks for the shout out! The book is a compilation of the course notes for
@StanfordPsych
Each chapter also has associated lecture slides that are available here:
Happy leaRning! 📈👍
Proud of what team MARPLE🕵️♀️has achieved in year 1!
We study how people use multi-modal evidence to figure out what happened and why, and develop computational models that can do so, too.
📎
🎬
The Moral Dynamics paper with
@flxsosa
,
@TomerUllman
, Josh Tenenbaum, and
@gershbrain
is now out in Cognition. Check it out!
📜preprint:
📰link:
📎repo:
🐦thread:
Excited to share Moral Dynamics, a model predicting moral judgments of agents in video clips based on intuitive physics and psychology! w/
@tobigerstenberg
@TomerUllman
@gershbrain
Josh Tenenbaum. Out as a preprint!
📄:
📎:
(1/7)
Excited to be holding the physical version of this paper in my hand. This was a surprise gift in my mailbox. Thanks to Jonathan Redshaw and Patricia Ganea for editing this special issue!
How do people make causal judgments? I argue that counterfactuals are key & hypotheticals don't suffice.
If you're like counterfactuals, hypotheticals, hipnoschweticals, what's the difference? 🙄This preprint is for you!
📰
📺
🧵👇
📣Now out in Cognition: "A counterfactual simulation model of causation by omission" with
@SimonStephan31
(we omitted the first 30 Simon Stephan's from the paper).
📰
📜
📎
1/8
📣 New paper from the CiCL! Zach Davis and
@KelseyRAllen
show that when a group of agents failed to coordinate, people can infer actions, capacities, and aspects of the situation from judgments of blame.
📰
📎
Wonderful talk by
@TamarKushnir
at
#CogSci2022
on how norms and obligations causally constrain actions, how the capacity to draw inferences from actions about mental states develops, and how this interacts with cultural experiences.
"Towards Singular Causal Explanations"
@ZennaTavares
shares his fascinating work with the Causality in Cognition Lab on marrying probabilistic programming languages with counterfactual inference to give birth to good explanations.
Check it out 👉
Brilliant psychology colloquium talk here
@Stanford
by
@fierycushman
telling us how we know what not to think, and that everything is possible (but only for philosophers).
How I'm joining the
#SPP2021
pre-conference workshop on Essentialism. On the
#Amtrak
from SF 🌴to Denver 🌄. Sacramento in the background and Susan Gelman () on the laptop.
GPT3's prediction was correct!
@LakeBrenden
's work presented in our Frisem
@StanfordPsych
is "shaking up the field". Brenden examines compositional generalization in humans, and builds machines that can do it, too. Thanks for a great talk!!
I'm super excited about this paper with
@Lara_Kirfel
and Thomas Icard. People infer event normality and causal structure from causal explanations 🤔💡
📰:
📎:
Thanks to
@jfkominsky
,
@xphilosopher
and
@phillipsjs
for feedback!
What do we learn from causal explanations?
In a new pre-print with Thomas Icard and
@tobigerstenberg
, we find that causal explanations disclose much more than what is explicitly stated - “Inference from explanation”:
A thread 🧵⬇️
#norms
#causality
Back from SPP
@socphilpsych
with many warm and fuzzy feelings 🧸❤️ What a wonderful meeting it was!
I'm already excited for the next one at Cornell University 🥳
I hope you were able to catch
@ErikBrockbank
and
@justintheyang
at the poster presentation session at
#CogSci2024
In "Without cookies, he's just a monster", they develop a counterfactual simulation model of how people explain behavior. w Mishika Govil &
@judyefan
🚫🍪🧌
Are LLMs more likely to categorize based on essential properties or described appearance?
@thesiyingzhg
, Jingyuan She & David Rose find that just like people, LLMs care more about what something is for than what it's made of.
📰
📎
Great talk by
@James__Dunlea
in our Developmental Psychology Brownbag on "Children's and adults' reasoning about punishment's messages". James' work unites social, moral, and dev psychology to answer questions at the heart of jurisprudential inquiry.
Thanks for joining us! 👍
What a wonderful
#SPP2021
keynote by Stanton prize winner
@phillipsjs
!! I can't wait to see all the future possibilities that will take shape on the white board behind him (it looks like there is still some space).
New preprint with
@SimonStephan31
"A counterfactual simulation model of causation by omission". When do people say that something happened because something else didn't happen?
📰
📎
🧵1/7
If you're headed to
#CDS2024
make sure to catch
@thesiyingzhg
's poster presentation on Saturday (P3-99). Siying will share a super cool new paradigm that probes children's counterfactual thinking without asking counterfactual questions! w David Rose, Sophie Bridgers &
@hyogweon
Brilliant colloquium talk
@StanfordPsych
by
@gershbrain
today on "How to never be wrong", arguing that robustness to disconfirmatory evidence can arise from purely rational principles.
@gershbrain
is never wrong!
📰
The WHY'21 workshop "Causal Inference & Machine Learning: Why now?" will take place this Monday at
#NeurIPS2021
. Our goal is to bring CI & ML researchers together to discuss the nextgen AI! Program (joint w/
@yudapearl
,
@bschoelkopf
, Y Bengio, T Sejnowski)
Today we had rockstar scientist (🚀⭐️👩🔬) Ishita Dasgupta () remotely present her work on "Causal reasoning from meta-reinforcement learning". Check it out here:
Fascinating talk today in our lab meeting by Megha Srivstava () on "Fairness and Robustness with Missing Information" showing how to build better ML algorithms by incorporating human causal knowledge.
📰