Publicly wondering if your product might be conscious is good hype @anthropic.com, but poor science.
Consciousness attribution should integrate prior expectations with public/shareable facts, not by computational equivalence alone or anecdotes. If that makes us "vulgar behaviorists", we are proud.
💡 Our new #preprint is available online!
How do people adapt their decisions when priorities change?
In our new study, we examine how the way people represent value shapes their ability to adjust in multi-goal environments.
🔗 OSF link: doi.org/10.31234/osf...
It's a thread 🧵
🧠🤖 New paper on AI & consciousness
We challenge the idea that computational equivalence is the right test for machine consciousness. In cognitive science, consciousness is never “read off” from code, it’s inferred inductively from behavior
with @thecharleywu.bsky.social
doi.org/10.1093/nc/n...
👇
Applications for BAMB! 2026 are officially open!
Join us in Barcelona (July 12–23) to master the art of behavioral modeling with our incredible faculty:
@meganakpeters.bsky.social
@marcelomattar.bsky.social
@khamascience.bsky.social
@thecharleywu.bsky.social
Apply now here: www.bambschool.org
This project was developed during my research visit to Prof. Benedetto De Martino’s lab at the UCL Institute of Cognitive Neuroscience @ucl.ac.uk, under the guidance of my supervisor Dr. Stefano Palminteri (@stepalminteri.bsky.social). I'm very grateful for this collaboration.
These findings suggest that how knowledge is structured can support behavioural flexibility when circumstances change.
Beyond human cognition, this question is also relevant for designing artificial systems that must operate under evolving objectives.
Across two experiments combining behavioural data and RL modelling, we found that:
👉 Participants using feature-based value representations adapted faster after goal switches
👉 Those relying on a single integrated value showed slower adjustment
We developed a multi-goal learning paradigm where participants had to infer changing priorities from aggregated outcomes.
This allowed us to test two possible adaptation strategies:
• Reweight previously learned features
• Relearn integrated option values
💡 Our new #preprint is available online!
How do people adapt their decisions when priorities change?
In our new study, we examine how the way people represent value shapes their ability to adjust in multi-goal environments.
🔗 OSF link: doi.org/10.31234/osf...
It's a thread 🧵
Moltbook, the AI social network, is a fascinating phenomenon.
Those interested in studying these kinds of systems may find our 2019 (!) Nature paper on 'Machine Behavior' useful.
www.nature.com/articles/s41...
🚨 New publication 🚨
Excited to share this new forum piece led by @kevinetiede.bsky.social
We propose broadening risk communication to cover neglected dimensions outlined in our risk-information taxonomy, requiring greater reliance on simulated experiences.
Open access: doi.org/10.1016/j.ti...
Across seven incentivized experiments and a large reanalysis, we systematically manipulated the presence and type of post-choice feedback in repeated risky decisions to test whether feedback shapes behavior through learning mechanisms or through anticipatory changes in preferences.
Across seven incentivized experiments and a large reanalysis, we systematically manipulated the presence and type of post-choice feedback in repeated risky decisions to test whether feedback shapes behavior through learning mechanisms or through anticipatory changes in preferences.
🚨 Deadline extension 🚨
BEHAVIORAL CLONES WORKSHOP
Imitating humans is a powerful route to machine behavior and an emerging lens on human cognition and social simulation.
🗓 Fri, 20 May 2026
📍 Following Machine+Behavior @ Max Planck Berlin
🗣️ Invited talks: Danica Dillion, Raja Marjieh, Marcel Binz
🧵 New paper in @NatureComms
Feedback-induced attitudinal changes in risk preferences
Nasioulas, Potier, Cerrotti, Lebreton & me (2026)
Does feedback really improve risky decision-making? Short answer: no! it changes attitudes, not learning. 👇
rdcu.be/e0VcO
Iranians are experiencing a collective trauma. Thousands have been killed/injured in recent events, the economy is crippled & the threat of a wider conflict is real. This is especially difficult for those living in Iran, as many have lost (or fear losing) loved ones. www.nature.com/articles/d41...
Calling all scientists: Support your Iranian colleagues www.nature.com/articles/d41...
Very happy that @PNASNews agreed to publish our (w/ @romanececchi.bsky.social) response to Prakhar's thought-provoking study! You can find the final version at the link below. See the following tweet for Prakhar's response to our response. Happy to hear your thoughts!
www.pnas.org/doi/10.1073/...
Our experiences have countless details, and it can be hard to know which matter.
How can we behave effectively in the future when, right now, we don't know what we'll need?
Out today in @nathumbehav.nature.com , @marcelomattar.bsky.social and I find that people solve this by using episodic memory.
🎉 My PhD work has just been published in @natcomms.nature.com!
How do we learn who caused what - and how much control we had - when outcomes depend on multiple people? We studied how humans do so using a new social learning task, computational modelling and fMRI.
www.nature.com/articles/s41...
🧵👇
Screenshot of the abstract: As large language models (LLMs) are increasingly integrated into decision-making systems—such as au-tonomous vehicles and medical devices—understanding how humans perceive and evaluate AI-generatedjudgements is crucial. To investigate this, we conducted a series of experiments in which participants evaluatedjustifications for moral and non-moral choices, generated either by humans or LLMs. Participants attempted toidentify the source of each justification (either human or LLM) and indicated their agreement with its content.We found that while detection accuracy was consistently above chance, it remained below 70%. In terms ofagreement, there was no overall preference for human-generated responses, even though machine-generatedjustifications were favored in particularly challenging moral scenarios. Notably, we observed a systematic anti-AIbias: participants were less likely to agree with judgments they believed were AI-generated, regardless of thetrue source. Linguistic cues, such as response length, typos, first-person pronouns, and cost-benefit languagemarkers (e.g., “lives,” “save”), influenced both detection and agreement. Participants tended to disagree withcost-benefit calculations, possibly due to an expectation that AI would favor such reasoning. These findingshighlight the influence of motivated belief and ingroup/outgroup bias in shaping human evaluation of AI-generatedcontent, particularly in morally sensitive contexts.
A revised version of our (@bsgarcia.bsky.social +Crystal Qian) paper “A Moral Turing Test to assess how subjective belief and objective source affect detection and agreement with LLM judgments” is now available on PsyArXiv
osf.io/preprints/ps...
With some trepidation, I'm putting this out into the world:
gershmanlab.com/textbook.html
It's a textbook called Computational Foundations of Cognitive Neuroscience, which I wrote for my class.
My hope is that this will be a living document, continuously improved as I get feedback.
Thanks Tanay 👾
Many thanks to my collaborators Martino Kuntze, Ilaria Dal Barco, Dr. Damien Challet, and my supervisor Dr. Stefano Palminteri @stepalminteri.bsky.social for their contribution and support.
I’m grateful to the Allianz team for highlighting early-career research at the intersection of climate risk, technology, and society.
Main result: resilience is not physical.
Small drops in cooperation or slower adaptation can push systems toward collapse, while strong norms help communities recover faster after shocks.
These insights provide signals like tipping points and time to recovery that complement climate risk models.
Using multi-agent reinforcement learning, we model two key cognitive factors:
• how strongly people follow shared rules (social norms)
• how quickly they adapt when conditions change
When droughts, heatwaves, or storms arrive, they do not
only strain resources, they stress the rules communities use to share them.
We study how community behavior affects resilience when shared resources face climate shocks.
🌊🎉 Our project has been selected among the Top 10 projects for the Allianz Climate Risk Award 2025 and is now featured in the official Allianz Climate Risk Award Compendium!
🔗 The fragile pact: resilience at the edge of scarcity (p.18)
www.allianz.com/content/dam/...
It's a thread 🧵
Excited to announce our symposium on how AI and humans shape each other
“Humans and Artificial Minds: Mutual Influences”
9 Jan at ENS Paris.
Free registration here www.eventbrite.com/e/humans-and...
Feel like this fresh year needs some serious culture change?
We got you covered! With @clarekelly.bsky.social @eikofried.bsky.social Anna van 't Veer
📝 rdcu.be/eXja4
