Finally, estimated trial-by-trial PEs from participants’ responses behaved similarly as insight experiences (Kullback-Leibler Divergence).
03.03.2026 15:47
👍 0
🔁 0
💬 0
📌 0
Finally, estimated trial-by-trial PEs from participants’ responses behaved similarly as insight experiences (Kullback-Leibler Divergence).
Indeed, in 3 datasets, the intensity of insight was best explained by an interaction between the accuracy and precision of initial predictions. More precision led to stronger insight for incorrect, but to less intense insight for correct predictions, corroborating ideas of insight as a readout of PE
We presented people ambiguous images for which they tried to guess the correct label (to derive prediction accuracy) and rated their confidence in that label (for prediction certainty). Insight intensity was rated as how strong subject’s “Aha!” was once the solution image was shown.
According to Bayesian inference, predictions are weighted by their respective uncertainty, influencing levels of surprise: violating more certain (vs. uncertain) predictions is more surprising.
We hypothesised a similar interaction for insight experiences if they are related to PEs.
Recent research proposed a relationship between insight experiences and the resolution of prediction errors (PEs) during sudden performance increases. So, more incorrect predictions, more intense insights?
Not quite! Bayesian inference highlights the role of the (un)certainty of predictions.
🚨Preprint! “Bayesian surprise tracks the strength of perceptual insight” - Work with @lindedomingo.bsky.social & @gonzalezgarcia.bsky.social
Ever wondered what factors influence the subjective experience of suddenly understanding a previously unclear input?
Click below:
doi.org/10.64898/202...
Yesterday at #ICON2025 I got to present our poster with #EttoreAmbrosini, @palencianoap.bsky.social, and @mruz.bsky.social on flexible task representations. I loved the chance to share our preliminary results and hear such thoughtful feedback!
We hope this work inspires new questions about this multifaceted phenomenon, which has received relatively little attention in the long-term memory field before.
Regarding long-term consequences, we predict a strong encoding of episodic and semantic information, supported by the integration of prediction errors and the semanticization of the input, but a weaker encoding of perceptual details.
We propose that the ambiguous stimulus initially induces prediction errors (PEs) across the hierarchy. Once disambiguation occurs, these PEs can be rapidly resolved and integrated, accompanied by a decrease in representational dimensionality as the disambiguated image is linked to existing schemas.
In this opinion paper, we explore the mnemonic consequences of sudden perceptual learning, i.e. the representational features characterising the abrupt disambiguation of an ambiguous stimulus (e.g., a Mooney image).
👩💻 New preprint out (FIRST one of my PhD journey) with the incredible @lindedomingo.bsky.social @ortiztudela.bsky.social @gonzalezgarcia.bsky.social
"From sudden perceptual learning to enduring engrams: A representational perspective" 🧠
doi.org/10.31234/osf...
🚨 There's an open postdoc position on sudden perceptual learning at the CIMCYC (University of Granada) - very nice people, excellent research environment, and Spanish lifestyle! Go check it out :)
Very excited to share our new preprint on visual ambiguity resolution. Check it out 👇
[🧵1/6] Would you like to know more about how the brain organizes complex, novel task information? Check out our last study in #JNeurosci @sfnjournals :
www.jneurosci.org/content/45/1...
@palencianoap.bsky.social @gonzalezgarcia.bsky.social @mruz.bsky.social @cimcyc.bsky.social
We will explore sudden learning in humans and artificial agents to uncover why these moments feel so sudden, which information we actually learn, and the role of prediction errors!
📣 ICON Symposium
Really excited to announce our symposium at @ICON this year on Sudden Learning Across Systems!
Together with lots of cool people: @gonzalezgarcia.bsky.social @lindedomingo.bsky.social @ortiztudela.bsky.social @anikaloewe.bsky.social @jazzmaniatico.bsky.social and Andrea Greve!
This is work together with @jvoeller.bsky.social, @vincentcheung.bsky.social, @numole.bsky.social, Konstantin Weise, Markus Kiefer and @gesahartwigsen.bsky.social at @unileipzig.bsky.social and @copla.bsky.social @mpicbs.bsky.social.
🚨 New preprint!! Using condition-and-perturb TMS, we show that functional interaction between multimodal and modality-specific cortices is causally relevant for conceptual knowledge retrieval: ssrn.com/abstract=510...