KInIT is a European Centre of Excellence in AI supported by

Research Group

Web & User Data Processing

On this page

Bio
Projects
Publications

Michal Spiegel

Research areas: machine learning, natural language processing, mechanistic interpretability, positional encodings, machine-generated text detection

Position: Volunteer

Email
LinkedIn
GitHub

Michal is a master’s student at Masaryk University, focusing on NLP and large language models. At KInIT, he authored numerous peer-reviewed papers (ACL 2024, SemEval 2024) and worked on deploying LLMs in scalable, cloud-native systems.

His current research explores algorithmic reasoning and positional encodings in transformers using the tools of mechanistic interpretability. Besides research, he has also found passion in teaching and popularising science and technology to the general public.

Selected Projects

VIGILANT: Vital IntelliGence to Investigate ILlegAl DisiNformaTion

VIGILANT is a Horizon Europe project aimed at application of data collection and AI-based web content analysis methods to support police authorities in investigations of disinformation-related crime. In this project,…

Selected Publications

IMGTB: A Framework for Machine-Generated Text Detection Benchmarking

Spiegel, M., Macko, D. – Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations),

Spiegel, M., Macko, D. In the era of large language models generating high quality texts, it is a necessity to develop methods for detection of machine-generated text to avoid their…

KInIT at SemEval-2024 Task 8: Fine-tuned LLMs for Multilingual Machine-Generated Text Detection

Spiegel, M., Macko, D. – Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024),

Spiegel, M., Macko, D. SemEval-2024 Task 8 is focused on multigenerator, multidomain, and multilingual black-box machine-generated text detection. Such a detection is important for preventing a potential misuse of large…