Lennart Luettgau

@LLuettgau

Research Scientist, AI Security, Societal Impacts @AISecurityInst. Cognitive neuroscientist by training.

Joined April 2019

338 Following 418 Followers 283 Posts

Latest posts 5

Lennart Luettgau @LLuettgau Sep 8, 2025

New preprint!
Growing numbers of people turn to AI chatbots for information, sparking debates about their potential to mislead voters and shape public discourse.
But what's the real impact of AI on political beliefs? Our latest research dives into this critical question 👇

Lennart Luettgau @LLuettgau Sep 1, 2025

My team at @AISecurityInst is hiring a Research Assistant to work on exciting Human Influence research projects!

-6 month residency
-Ideal for recent MSc or early PhD students in ML, AI, Psych, Cognitive/computer/data science

(link in thread)

Lennart Luettgau @LLuettgau May 12, 2025

Excited to share some work recent work on using hierarchical Bayesian modeling to improve LLM evaluations! HiBayES is a flexible and robust statistical modeling framework that allows to explicitly account for the hierarchical nature of benchmarks

AI Security Institute (AISI) Verified account @AISecurityInst May 12, 2025

Advanced AI systems require complex evaluations to measure abilities, but conventional analysis techniques often fall short.
Introducing HiBayES: a flexible, robust statistical modelling framework that accounts for the nuances & hierarchical structure of advanced evaluations.

View quoted post on X

Lennart Luettgau @LLuettgau Nov 11, 2024

Excited to announce that this work has now been published in @PNASNews

pnas.org/doi/10.1073/pn…

We propose a computational account of how humans decompose complex dynamical experiences into subprocesses and reuse these in an abstract, compositional manner in novel situations.

www.pnas.org Decomposing dynamical subprocesses for compositional generalization | PNAS A striking feature of human cognition is an exceptional ability to rapidly adapt to novel situations. It is proposed this relies on abstracting and...

Lennart Luettgau @LLuettgau Sep 3, 2023

As promised, here’s the summary thread🧵:
The study presented in this preprint investigates how humans decompose dynamical subprocesses and leverage the abstracted subprocesses for compositional reuse of experience in new situations.
1/n
View quoted post on X

Lennart Luettgau @LLuettgau Sep 23, 2024

Excited to share our brand-new work shedding some light on the neural mechanisms behind one of human’s coolest cognitive feats: compositional generalization of structural knowledge!
A Tweeprint-Thread 🧵
1/n