Excited about our new work measuring multi-turn persuasion in AI-human interactions and how to simulate human persuadability!

LLMs can shift people's beliefs. But most persuasion studies only check beliefs before and after a conversation. We built PersuasionTrace to measure beliefs turn by turn, so we can study how belief updates actually unfold.

Task diversity is supposedly key to generalization in RL. But what does it do to continual RL, where agents face one new task distribution after another? We find that past a point, more diversity actually inhibits continual reinforcement learning 🧵

Really excited to have the opportunity to give a talk on this work @cogscisociety.bsky.social !!! last year was a blast can’t wait to go back to Rio in July 🇧🇷 HUGE thanks to my collaborators for the support @aydanhuang265.bsky.social @EricYe29011995 @natashajaques.bsky.social @maxkw.bsky.social 🙏

Our new short piece in TiCS on intuitive theories of truth: how people judge whether statements could be true, whether statements are true, and whether to assert them as true. A great collab with @keremoktar.bsky.social @ihandleyminer.bsky.social @kevinzollman.com @lianeleeyoung.bsky.social

Can't wait to present this work @iclr-conf.bsky.social this year!!! Looking forward to hearing everyone's thoughts on the paper and learning more about peoples' research! Thanks again to my collaborators for all of their help on this project!

🤔💭What even is reasoning? It's time to answer the hard questions! We built the first unified taxonomy of 28 cognitive elements underlying reasoning Spoiler—LLMs commonly employ sequential reasoning, rarely self-awareness, and often fail to use correct reasoning structures🧠

Forget modeling every belief and goal! What if we represented people as following simple scripts instead (i.e "cross the crosswalk")? Our new paper shows AI which models others’ minds as Python code 💻 can quickly and accurately predict human behavior! shorturl.at/siUYI%F0%9F%...

New paper challenges how we think about Theory of Mind. What if we model others as executing simple behavioral scripts rather than reasoning about complex mental states? Our algorithm, ROTE (Representing Others' Trajectories as Executables), treats behavior prediction as program synthesis.

When values collide, what do LLMs choose? In our new paper, "Generative Value Conflicts Reveal LLM Priorities," we generate scenarios where values are traded off against each other. We find models prioritize "protective" values in multiple-choice, but shift toward "personal" values when interacting.

Excited by our new work estimating the empowerment of LLM-based agents in text and code. Empowerment is the causal influence an agent has over its environment and measures an agent's capabilities without requiring knowledge of its goals or intentions.

Claire's new work showing that when an assistant aims to optimize another's empowerment, it can lead to others being disempowered (both as a side effect and as an intentional outcome)!

Still catching up on my notes after my first #cogsci2025, but I'm so grateful for all the conversations and new friends and connections! I presented my poster "When Empowerment Disempowers" -- if we didn't get the chance to chat or you would like to chat more, please reach out!

lol this may be the most cogsci cogsci slide I've ever seen, from @maxkw.bsky.social "before I got married I had six theories about raising children, now I have six kids and no theories"......but here's another theory #cogsci2025

Our new paper is out in PNAS: "Evolving general cooperation with a Bayesian theory of mind"! Humans are the ultimate cooperators. We coordinate on a scale and scope no other species (nor AI) can match. What makes this possible? 🧵 www.pnas.org/doi/10.1073/...

As always, CogSci has a fantastic lineup of workshops this year. An embarrassment of riches! Still deciding which to pick? If you are interested in building computational models of social cognition, I hope you consider joining @maxkw.bsky.social, @dae.bsky.social, and me for a crash course on memo!

Very excited for this workshop!

#Workshop at #CogSci2025 Building computational models of social cognition in memo 🗓️ Wednesday, July 30 📍 Pacifica I - 8:30-10:00 🗣️ Kartik Chandra, Sean Dae Houlihan, and Max Kleiman-Weiner 🧑‍💻 underline.io/events/489/s...

'Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination' @kjha02.bsky.social · Wilka Carvalho · Yancheng Liang · Simon Du · @maxkw.bsky.social · @natashajaques.bsky.social doi.org/10.48550/arX... (3/20)

Settling in for my flight and apparently A.I. DOOM is now a movie genre between Harry Potter and Classics. Nothing better than an existential crisis with pretzels and a ginger ale.

Thanks to the Diverse Intelligence Community for all these inspiring days & impressions in Sydney 🙏🏻 @chriskrupenye.bsky.social @katelaskowski.bsky.social @divintelligence.bsky.social @maxkw.bsky.social

LLMs learn beliefs and values from human data, influence our opinions, and then reabsorb those influenced beliefs, feeding them back to users again and again. We call this the "Lock-In Hypothesis" and develop theory, simulations, and empirics to test it in our latest ICML paper!

Excited to speak about some new work on Bayesian Cooperation at this workshop! Join us virtually

Now out in JPSP ‼️ "Inference from social evaluation" with Zach Davis, Kelsey Allen, @maxkw.bsky.social, and @julianje.bsky.social 📃 (paper): psycnet.apa.org/record/2026-... 📜 (preprint): osf.io/preprints/ps...

Our new paper (first one of my PhD!) on cooperative AI reveals a surprising insight: Environment Diversity > Partner Diversity. Agents trained in self-play across many environments learn cooperative norms that transfer to humans on novel tasks. shorturl.at/fqsNN%F0%9F%...

Awesome new work from my lab led by @kjha02.bsky.social scaling cooperative AI! True cooperation requires adapting to both unfamiliar partners and novel environments. Agents trained with CEC get us closer to agents that can act with general cooperative principles rather than memorized strategies.

How AlphaGo like architectures can explain human insight. Out now in Cognition!

my paper with max, @maxkw.bsky.social, tuomas, and @fierycushman.bsky.social out in cognition at long last www.sciencedirect.com/science/arti... We explain why humans and successful AI planners both fail on a certain kind of problem that we might describe as requiring insight or creativity

Accepted as a Spotlight in ICLR2025!

Emergent transition from code to natural language for reasoning tasks when RL tuning a language model for math. Interesting to consider implications for "Language of Thought" style theories in cognition. hkust-nlp.notion.site/simplerl-rea...

🔊 New paper just accepted in JPSP 🥳 In "Inference from social evaluation", we explore how people use social evaluations, such as judgments of blame or praise, to figure out what happened. 📜 osf.io/preprints/ps... 📎 github.com/cicl-stanfor... 1/6

Very nice to see our work on LLM agent cooperation covered in Wired! www.wired.com/story/ai-soc...

Josh Tenenbaum on scaling up vs growing up and the path to human-like reasoning #NeurIPS2024

Honored to receive the Best Paper Award at the #NeurIPS2024 Pluralistic Alignment Workshop Check out our preprint "Language Model Alignment in Multilingual Trolley Problems" at arxiv.org/pdf/2407.02273!

Really excited to present my work this Sunday @NeurIPS on how we might approach training a generalist agent capable of cooperation at scale: coordinating with many novel partners on many novel tasks has never been easier! Come by the IMOL workshop to check it out and chat more!

Come hang!

Now with a luma link: lu.ma/1bxqun5l

Only at NeurIPS

Excited that our multi-agent LLM agent work, “Cooperate or Collapse: Emergence of Sustainable Cooperation in a Society of LLM Agents,” will be presented at #NeurIPS24 -- reach out if you want to meet up in Vancouver!

Github Copilot output. A sad but fascinating alignment failure. Reveals hidden LLM biases by going out of their RLHF distribution

Overparenting is on the rise and hurts children’s motivation starting in early childhood. How can we help parents step back? Our new paper in Child Dev shows that pointing out learning opportunities reduces overparenting. srcd.onlinelibrary.wiley.com/doi/10.1111/...

The computational cognitive science starter kit

I’m recruiting PhD students to join the Computational Minds and Machines Lab at the University of Washington in Seattle! Join us to work at the intersection of computational cognitive science and AI with a broad focus on social intelligence. (Please reshare!)

Got it started go.bsky.app/9gsefkW

Okay the people requested one so here is an attempt at a Computational Cognitive Science starter pack -- with apologies to everyone I've missed! LMK if there's anyone I should add! go.bsky.app/KDTg6pv

These videos of a humanoid watching itself in a mirror are a cool test of a “self” model. The robot’s world model tries to predict the next visual inputs. It can predict its own movements from a first-person view but doesn’t map the mirrored body to its own, and thus, the

These videos of a humanoid watching itself in a mirror are a cool test of a “self” model. The robot’s world model tries to predict the next visual inputs. It can predict its movements from a first-person view but doesn’t map the mirrored body to its own, and thus, the mirrored view doesn't match.

Many point to cumulative culture as the distinctive feature of human intelligence. Now, a new theory argues that our success comes from the way that cumulative culture is uniquely open-ended, i.e., we can accumulate novel technologies, languages, moral systems, or nearly anything

Many point to cumulative culture as the distinctive feature of human intelligence. Now, a new theory argues that our success comes from the way that cumulative culture is uniquely open-ended, i.e., we can accumulate novel technologies, languages, moral systems, or nearly anything adaptive.

Incredible to see that the fruit fly connectome has been completed. I remember as an undergraduate watching Sebastian Seung in 2007 give the presidential lecture at SFN mysteriously titled: “The Once and Future Science of Neural Networks.”