Sign in

kenzakdr.bsky.social

@kenzakdr.bsky.social
33 followers 30 following 6 posts
PostsRepliesMedia
Reposted by @kenzakdr.bsky.social
Taiki Oka @toka-neuro.bsky.social · 03/09/2026
1/7 Excited to share our new @elife.bsky.social Reviewed Preprint! 🧠 How do higher-order cognitive processes like abstraction & metacognition map onto transdiagnostic mental health dimensions? We combined computational modeling of a RL task with symptom profiling 🔗 doi.org/10.7554/eLif...
doi.org
Mapping abstraction and metacognition onto distinct transdiagnostic symptom profiles
11610
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
6/6 Why it matters: biased information gathering is a hallmark of conditions like schizophrenia and OCD. Mapping these separable prefrontal-subcortical computations gives a mechanistic handle on where sampling can go awry. w/ @aleyamarzuki.bsky.social Magdalena del Rio @tobiasuhauser.bsky.social
031
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
5/6 Pulling that signal apart, the individual processes turn out to live in distinct circuits: urgency tracked in locus coeruleus + dmPFC, accumulated evidence (ESd-1) in anterior-medial PFC, and evidence-strength prediction errors (ΔES) in ventral striatum + dmPFC.
120
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
4/6 The key neural result: we found an integrated "keep sampling vs. stop" signal spanning ventromedial and dorsomedial PFC, a hub that seems to pool the different ingredients of the decision into one overall drive to keep gathering info.
121
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
3/6 Behaviourally, Behaviourally, two things stood out: people showed escalating urgency to stop sampling as trials went on, and a systematic bias in how they weighted information over time: recent evidence was weighted more than older evidence.
110
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
2/6 Participants did a sequential information-gathering task during fMRI. On each trial, they saw draws of 5 coloured cues (2 possible colours) and could keep drawing to gather more evidence before deciding which colour was most common. Wait too long, though, and the trial timed out.
110
kenzakdr.bsky.social @kenzakdr.bsky.social · 18/08/2026
1/6 New preprint! 🧵 How do we decide to decide? With @aleyamarzuki.bsky.social we showed that the interplay between urgency, evidence accumulation, and prediction errors converges in the dmPFC, into one signal: keep gathering information, or stop and commit to a choice. doi.org/10.64898/202...
doi.org
1178
Reposted by @kenzakdr.bsky.social
Leya @aleyamarzuki.bsky.social · 14/08/2026
📢 Our paper is now out in @natcomms.nature.com! www.nature.com/articles/s41... We know that symptoms of OCD fluctuate substantially. But what are the exact conditions that precede symptom escalations? We found evidence for metacognitive processes being a key antecedent. 1/
nature.com
Drops in confidence precede symptoms of obsessive-compulsive disorder - Nature Communications
To devise effective interventions for obsessive-compulsive disorder, it is crucial to understand antecedents to symptom emergence. Here, the authors show evidence for self- and decision making underco...
12512
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 07/08/2026
🧠New preprint: OSCAR, an auto-associative model that generates its own error signal by reconstructing its input, and so fits the inverse base-rate effect across supervised, observational, and unsupervised paradigms with one single architecture. Two new experiments included. arxiv.org/abs/2608.06149
arxiv.org
Errorless Irrationality: A unified computational account of the inverse base-rate effect across predictive, observational, and unsupervised procedures
The inverse base-rate effect is a robust bias in how people resolve ambiguity between competing categories, and the most prominent theories explain it through prediction error. Across two experiments ...
143
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 14/07/2026
📢 Out now in PLOS Comp Biol: **cpm** — a Python library for theory-driven modelling in computational psychiatry. No scikit-learn for mechanistic cognitive models — so we built one: RL, prospect-theory & metacognition, hierarchical Bayes, and as many more as you can code 🔗 doi.org/10.1371/jour...
doi.org
cpm: A python library for theory-driven modelling in computational psychiatry
Author summary Computational psychiatry is a field that makes extensive use of computational methods to understand mental health. In this field and related disciplines, there is a strong need to desig...
12813
Reposted by @kenzakdr.bsky.social
Ben Jonathan Wagner @benjwagner.bsky.social · 30/05/2026
We @saschafrolich.bsky.social @stefankiebel.bsky.social investigated how habitual- and goal-directed control interact when actions must be made under time pressure. www.nature.com/articles/s44...
nature.com
Context-dependent interaction between goal-directed and habitual control under time pressure - Communications Psychology
This study investigates the dynamic interplay between habitual and goal-directed control. Using RL and drift-diffusion modelling, we show that these systems interact in a context-dependent manner: hab...
11614
Reposted by @kenzakdr.bsky.social
Tricia Seow @seowxft.bsky.social · 12/11/2025
Very pleased to share our new review in @natrevpsychol.nature.com on integrating cognitive neuroscience and clinical perspectives of metacognition in mental health! With @tobiasuhauser.bsky.social, Lena Jelinek, and Steffen Moritz. (1/3) Web: www.nature.com/articles/s44... PDF: rdcu.be/ePxUG
eur01.safelinks.protection.outlook.com
Integrating cognitive neuroscience and clinical perspectives on metacognitive mechanisms in psychopathology - Nature Reviews Psychology
Cognitive neuroscience and clinical psychology have made substantial advances in knowledge about metacognitive processes, but these fields have progressed in parallel. In this Review, Seow et al. inte...
34313
Reposted by @kenzakdr.bsky.social
Tobias Hauser @tobiasuhauser.bsky.social · 09/10/2025
Come and say hi to @seowxft.bsky.social @floorburghoorn.bsky.social @lenarddome.bsky.social & @kenzakdr.bsky.social who will be demonstrating our brain explorer app and research at the @medresfdn.bsky.social stand at New Scientist Live event in London @uclqsion.bsky.social @unituebingen.bsky.social
086
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 16/09/2025
A new preprint 📝 with @tobiasuhauser.bsky.social @kenzakdr.bsky.social @benjwagner.bsky.social and Andrew Webb accompanying our cpm-toolbox.net python modelling library - including details about our motivations, toolbox features, framework and workflows! 👉 osf.io/preprints/ps...
084
Reposted by @kenzakdr.bsky.social
Leya @aleyamarzuki.bsky.social · 11/09/2025
🚨New pre-print out! 🍂 Using daily gamified smartphone 📱 tracking, we found that reductions in both self- and decision-making confidence precede when people are likely to report experiencing symptoms of OCD. www.researchsquare.com/article/rs-7...
researchsquare.com
Metacognitive antecedents to states of mental ill-health: Drops in confidence precede symptoms of OCD
Mental health symptoms, like those in obsessive-compulsive disorder (OCD), show pronounced fluctuations over time. However, little is known about the underlying factors driving these fluctuations. Whi...
23112
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 18/07/2025
A quote from Rumelhart/McClelland: "...real biological systems cannot be Turing machines because they have finite hardware." A sentiment somewhat related to our g-distance framework. What portion of unobserved but possible results your model rejects?
141
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 14/07/2025
Come check out my poster on an accessible new python library cpm (cpm-toolbox.net) designed to ease you into modelling in computational psychiatry #CPConf2025 I’m poster 1.2 👀
073
Reposted by @kenzakdr.bsky.social
Lénárd Döme @lenarddome.bsky.social · 16/04/2025
This one is finally out! 🎉 We introduce g-distance, a measure that assesses model adequacy by comparing the range of behaviours exhibited by models to that of humans. #computationalmodeling #cognitivescience psycnet.apa.org/doi/10.1037/...
11710