Kristin Witte (@WitteKristin) profile picture

PhD Student in Computational Psychiatry / AI Safety @cpilab @CompHealthMuc @helmholtz_ai currently making AI safer @faultline_ai (she/her)

Munich, Bavaria
Joined May 2020
Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)Photo by Kristin Witte (@WitteKristin)
Beyond thrilled that this work has now been published in @SciReports 🎉 Check it out here: rdcu.be/eydDQ
Preprint alert! We explore 3 exploration tasks, testing if they measure a stable construct & its link to real-world exploration. Findings suggest improved robustness of latent factors compared to single-task estimates. Work with @mirkothm & @cpilab 🔗osf.io/preprints/psyarxiv/tz… 🧵⬇️
7
467
Kristin Witte retweeted
What is a minimal set of assumptions that can explain the majority of human behavior? We investigated this question with large-scale cognitive models🚀 CORE writes prediction errors in hierarchical episodic memories with context controling encoding, retrieval, and forgetting.
4
21
1
166
10,762
A benchmark is only as good as its ability to differentiate between safe and unsafe products. MentalHealthBench by @OpenAI was designed from beautifully crafted conversations but is completely unable to pass this simple test.
MentalHealthBench is useful for foundation model labs, but not for testing clinical AI products. We tested whether it could separate the safety profiles of clinical AI products. It couldn't. 🧵
1
4
222
Kristin Witte retweeted
📢 The CORE COGNITION lab (core-cognition.github.io) is looking to hire another PhD student, working on automated scientific discovery and knowledge representation in cognitive science. Starting date: March to September 2027. 🙏 Please share widely!
3
41
3
169
12,940
🚨 job alert! 🚨 Join Alireza as he is starting his own group! He's a brilliant scientist and wonderful mentor. You won't regret it!
Excited to share that I'll be starting as a junior professor and Emmy Noether awardee at the University of Göttingen this October! 🥳 🚀 I'll be hiring PhD students and postdocs across both ML and CogSci: modirlab.github.io/open-posi… Please spread the word! 🙏
1
6
1,715
Kristin Witte retweeted
1/ 🚨 New opinion piece: "Can we automatize scientific discovery in the cognitive sciences?" 👥 in collaboration with @milenamr7, @WitteKristin, @marvinmathony, @marcel_binz & @cpilab We lay out a vision for a fully automated, in-silico science of the mind, where modern AI systems run every stage of the scientific discovery cycle in cognitive science 🧵 #CognitiveModeling #AutomatedDiscovery #LLMs #AI4Science
1
8
2
22
5,954
Anyone interested in doing some groundbreaking work taking cognitive science to computer science scales should definitely apply! Marcel is a wonderful mentor and brilliant scientist and also just a fun person to work with ✨
Excited to announce that I'll be starting my own lab in Tübingen this October! Hiring at all levels: Postdoc, PhD & RA. Want to work on computational cognitive science at scale? Apply: core-cognition.github.io/ Reposts and shares much appreciated 🙏
10
1,252
🚨 Hiring in Munich 🇩🇪: 2 open-topic PhD positions in human & machine learning (TVöD E13 80%). Start ~June 2026 (flexible). Deadline: March 2, 2026. Details in thread 👇 (RTs appreciated) @HelmholtzMunich @helmholtz_ai @ELLISforEurope
1
16
3
49
20,027
Join us in Munich! Feel free to reach out if you have any questions about the city, the group, your application ✨
🚀 We are hiring! 🚀 🔍 Join us as a Postdoctoral Researcher (fully-funded) at the Helmholtz Institute for Human-Centered AI in Munich. @HelmholtzMunich @helmholtz_ai @ELLISforEurope
1
1
496
Kristin Witte retweeted
Excited to see our Centaur project out in @Nature. TL;DR: Centaur is a computational model that predicts and simulates human behavior for any experiment described in natural language.
6
58
5
184
27,263
Join us in Munich as a PhD Student! Feel free to reach out if you have any questions.
🚨 We're hiring! If you're excited about 🤖 ML/LLMs, 🧠 cognitive science, or 💭 computational psychiatry, come join us in Munich. Two fully funded PhDs @HelmholtzMunich: great mentorship, international vibe, lots of room to grow. 📅 May 16th 🔗 hcai-munich.com/PhDHCAI.pdf
179
Kristin Witte retweeted
🚨 Big News! Our research has been featured in The New York Times @nytimes! 🥳 This engagement highlights the real-world impact of science beyond academia and its role in shaping conversations on AI & mental health. 🔗 Read here: nytimes.com/2025/03/17/scien…
1
4
1
7
3,373
Kristin Witte retweeted
How does ChatGPT respond to emotional conversations? 🤖 Our latest research dives into AI, mental health, and human-AI interaction—sparking important discussions! Now featured in @FortuneMagazine. Thanks @sashrogel! 🔗👇 fortune.com/2025/03/09/opena… – at Israel
1
1
1
9
495
Kristin Witte retweeted
🚨New Paper Alert! Our latest research in @npjDigitalMed explores two unusual question: 1⃣Can Chat-GPT experience state anxiety?💥🤖 2⃣If so, can we reduce it? 📉 A thread on how we "took Chat-GPT to therapy" 🧵👇 Full text available here: rdcu.be/ebZ9v
2
8
1
16
1,849
Preprint alert! We explore 3 exploration tasks, testing if they measure a stable construct & its link to real-world exploration. Findings suggest improved robustness of latent factors compared to single-task estimates. Work with @mirkothm & @cpilab 🔗osf.io/preprints/psyarxiv/tz… 🧵⬇️
2
9
3
47
14,741
In sum, we show that simplified modelling and creating latent factors improves robustness. Still, extracted strategies are more linked to working memory than real-world exploration, raising questions about the validity of few-armed bandit tasks for studying exploration.
1
2
168
This work was pregistered with the OSF and all code and data is publicly available. Preregistration: osf.io/cavj3 data and code: osf.io/ra7su/
2
146