PhD student in Computational Cognitive Sciences | ENS-PSL. Currently working on the dynamics of impression formation, reputation management, and how it impacts our behavior.
https://mariusmercier.github.io
Marius Mercier
Loading...
Happy to have presented at @culturalevolsoc.bsky.social the work we did with Mia Karabegovic & @hugoreasoning.bsky.social on the moral psychology of intellectual property : moral intuitions about IP are less about protecting ideas, and more about protecting accurate reputation attribution.
New paper accepted as a proceedings of the Cognitive Science Society: "Inferring arithmetic skills from speed and accuracy”!
We tested whether people were optimal in their inference of others' math skills.
doi.org/10.31234/osf...
Next week, I will be presenting this work and some extensions at #HBES2026 in Rabat, Morocco.
Happy to connect!
In a second study, we wanted to test whether participants could integrate speed with accuracy cues. We provided them with additional information that the other person solved the question under a time limit of 20s (Fast condition) or 40s (Slow condition, within-subjects).
Participants had to predict the questions someone could solve, given the information that this person solved (or not) another question.
Their predictions were highly accurate compared to our benchmark data and matched those of an optimal Bayesian model. Even participants bad at math were accurate!
Huge thanks to my co-authors Ruby de Lanerolle (co-first author), @oliviermorin.bsky.social, @tadegquillien.bsky.social and to my supervisor @hugoreasoning.bsky.social
I will be presenting the paper in Rio for CogSci2026, happy to chat :)
doi.org/10.31234/osf...
Participants were still accurate in their prediction, but they did not attribute higher competence in the Fast condition.
One possible explanation is that the time limit was not diagnostic enough to change participants’ predictions (they were already accurate).
Curious to hear what you think!
Grégoire Darcy
Marius Mercier
Marius Mercier
Our new paper with @maxtaylordavies.bsky.social introduces a resource-rational model of Theory of Mind.
The model can explain many of the successes and failures of mindreading in human adults and children, and non-human primates. 🧵
royalsocietypublishing.org/rspb/article...
