iiinova — where your digital self keeps evolving

Reposted by Maëva L'Hôtellier

Constance Destais @constancedestais.bsky.social · 10/03/2026

I’m delighted to be speaking at Semaine du Cerveau in Paris about the science of everyday decision-making — the hidden mechanisms shaping how we think, choose, and learn, from confirmation bias to confidence. 🧠 Join us Wed March 18 18:30 at École normale supérieure ! tinyurl.com/ywamh75a

semaineducerveau.fr

Libres ou biaisé·es? La science des décisions du quotidien - Semaine du Cerveau

Entre liberté et automatismes, nos décisions du quotidien sont souvent influencées par une force discrète : le biais de confirmation. Autrement dit, nous sommes plus sensibles à ce qui nous donne rais...

0103

Reposted by Maëva L'Hôtellier

Ali Shiravand @alishiravand.bsky.social · 09/02/2026

💡 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 🧵

13011

Reposted by Maëva L'Hôtellier

Caroline Pioger @carolinepioger.bsky.social · 17/11/2025

I recently attended the Center for Decision Sciences Summer School at Royal Holloway 🇬🇧 a fantastic week exploring computational approaches to decision-making 🤖🧠 I wrote a recap of what we learned each day, for anyone curious about the content or experience ✨ center-decision-sciences.com/feedback/

center-decision-sciences.com

Feedback

Program Registration Schedule Directions Survey Responses Home Blog written by Caroline Pioger PhD student at Ecole Normale Supérieure, Paris, France The first edition of the Center for Decision Sc…

101

Reposted by Maëva L'Hôtellier

Ali Shiravand @alishiravand.bsky.social · 16/11/2025

🧠Our new preprint is out on PsyArXiv! We study how getting more feedback (seeing what you could have earned) and facing gains vs losses change the way people choose between risky and safe options. 🖇️Link: doi.org/10.31234/osf... It's a thread🧶:

13110

Reposted by Maëva L'Hôtellier

Steven @stevengeysen.bsky.social · 22/10/2025

‼️New preprint‼️ There does not seem to be an effect of ghrelin on risky decision-making in probability discounting. Not in behaviour, underlying computational processes, or neural activity. More details ⬇️

283

Reposted by Maëva L'Hôtellier

bioRxiv Neuroscience @biorxiv-neursci.bsky.social · 28/06/2025

Interoception vs. Exteroception: Cardiac interoception competes with tactile perception, yet also facilitates self-relevance encoding www.biorxiv.org/content/10.1101/202…

01511

Reposted by Maëva L'Hôtellier

Stefano Palminteri @stepalminteri.bsky.social · 11/06/2025

Lucky for you, lazy people at #RLDM2025, two of the best posters have apparently been put side-by-side: go check @maevalhotellier.bsky.social and @constancedestais.bsky.social posters!

0304

Reposted by Maëva L'Hôtellier

Romane Cecchi @romanececchi.bsky.social · 22/04/2025

🧵 New preprint out! 📄 "Elucidating attentional mechanisms underlying value normalization in human reinforcement learning" 👁️ We show that visual attention during learning causally shapes how values are encoded w/ @sgluth.bsky.social & @stepalminteri.bsky.social 🔗 doi.org/10.31234/osf...

doi.org

OSF

1176

Reposted by Maëva L'Hôtellier

fabiencerrotti.bsky.social @fabiencerrotti.bsky.social · 16/04/2025

🚨 New preprint on bioRxiv! We investigated how the brain supports forward planning & structure learning during multi-step decision-making using fMRI 🧠 With A. Salvador, S. Hamroun, @mael-lebreton.bsky.social & @stepalminteri.bsky.social 📄 Preprint: submit.biorxiv.org/submission/p...

submit.biorxiv.org

bioRxiv Manuscript Processing System

Manuscript Processing System for bioRxiv.

2239

Reposted by Maëva L'Hôtellier

Ali Shiravand @alishiravand.bsky.social · 16/02/2025

🎉 I'm excited to share that 2 of our papers got accepted to #RLDM2025! 📄 NORMARL: A multi-agent RL framework for adaptive social norms & sustainability. 📄 Selective Attention: When attention helps vs. hinders learning under uncertainty. Grateful to my amazing co-authors! *-*

1157

Reposted by Maëva L'Hôtellier

Charley Wu @thecharleywu.bsky.social · 10/02/2025

🚨 Finally out! My new @annualreviews.bsky.social in Psychology paper: www.annualreviews.org/content/jour... We unpack why psych theories of generalization keep cycling from rigid rule-based models to flexible similarity-based ones, then culminating in Bayesian hybrids. Let's break it down 👉 🧵

411942

Reposted by Maëva L'Hôtellier

Stefano Palminteri @stepalminteri.bsky.social · 23/01/2025

Epistemic biases in human reinforcement learning: behavioral evidence, computational characterization, normative status and possible applications. A quite self-centered review, but with a broad introduction and conclusions and very cool figures. Few main takes will follow osf.io/preprints/ps...

13819

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Link to the preprint: osf.io/preprints/ps...

osf.io

OSF

020

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Questions or thoughts? Let’s discuss! Reach out — we’d love to hear from you! 🙌

100

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Why does it matter? 🤔 Our work aims at bridging cognitive science and machine learning, showing how human-inspired principles like reward normalization can improve reinforcement learning AI systems!

120

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

What about Deep Decision Trees? 🌳 We further extend the RA model by integrating a temporal difference component to the dynamic range updates. With this extension, we demonstrate that the magnitude invariance capabilities of the RA model persist in multi-step tasks.

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

With this enhanced model, we generalize the main findings to other bandit settings: The dynamic RA model outperforms the ABS model in several bandit tasks with noisy outcomes, non-stationary rewards, and even multiple options.

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Once these basic properties are demonstrated in a simplified set-up, we enhance the RA model to successfully cope with stochastic and volatile environments, by dynamically adjusting its internal range variables (Rmax / Rmin).

120

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

In contrast, the RA model, by constraining all rewards to a similar scale, efficiently balances exploration and exploitation without the need for task-specific adjustment!

121

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Crucially, modifying the value of the temperature (𝛽) from the Softmax function does not solve the problem of the standard model. It simply shifts the peak performance along the magnitude axis. Thus, to achieve high performance, the ABS model requires tuning the 𝛽 value to the magnitudes at stake.

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Agent-Level Insights: ABS performance drops to chance due to over-exploration in small rewards and over-exploitation in large rewards. In contrast, the RA model maintains a consistent, scale-invariant performance.

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

First, we simulate ABS and RA behavior in bandits tasks with various magnitude and discriminability levels. As expected the standard model is highly dependent on the tasks levels, while the RA model achieves high accuracy over the whole range of values tested!

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

To avoid magnitude-dependence, we propose the Range-Adapted (RA) model: RA normalizes rewards, enabling consistent representation of subjective values within a constrained space, independent of reward magnitude.

111

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

Standard reinforcement learning algorithms encode rewards in an unbiased, absolute manner (ABS), which make their performance magnitude-dependent.

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

This work was done in collaboration with Jérémy Pérez, under the supervision of @stepalminteri.bsky.social 👥 Let's now dive into the study!

110

Maëva L'Hôtellier @maevalhotellier.bsky.social · 10/12/2024

New preprint! 🚨 Performance of standard reinforcement learning (RL) algorithms depends on the scale of the rewards they aim to maximize. Inspired by human cognitive processes, we leverage a cognitive bias to develop scale-invariant RL algorithms: reward range normalization. Curious? Have a read!👇

1147

Reposted by Maëva L'Hôtellier

Stefano Palminteri @stepalminteri.bsky.social · 05/12/2024

🚨New preprint alert!🚨 Achieving Scale-Invariant Reinforcement Learning Performance with Reward Range Normalization. Where we show that things we discover in psychology can be useful for machine learning. By the amazing @maevalhotellier.bsky.social and Jeremy Perez. doi.org/10.31234/osf...

osf.io

OSF

12310