21.10.2025
Emil Partow Takes Second Place in ESA Kaggle Competition
Detecting Manipulated LLM Outputs
MCML Junior Member Emil Partow achieved second place in the European Space Agency's Kaggle competition "Fake or Real: The Impostor Hunt in Texts." The challenge brought together participants to develop methods for identifying manipulated outputs from large language models.
The competition addresses key AI security risks, including data poisoning and overreliance on large language models. Participants were tasked with distinguishing outputs from correctly functioning models from those produced by maliciously modified LLMs, contributing to the development of more secure and trustworthy AI systems for space operations and beyond.
Congratulations on this outstanding achievement!
Related
12.11.2026
One Toolbox for Many Kinds of Uncertainty
MCML researchers introduce probly, a unified Python package for uncertainty-aware machine learning, presented orally at NeurIPS 2026.
09.11.2026
Outstanding Paper Award at EMNLP 2026 for Work on Accent Bias in Audio Language Models
A paper co-authored by MCML PI Valentin Hofmann on accent bias in audio language models won an Outstanding Paper Award at EMNLP 2026.
