Petter Törnberg

A computational model showing how clustered online networks can accelerate the spread of misinformation as a complex contagion process.

Shows that digital media often intensify polarization through partisan sorting dynamics rather than direct persuasion alone.

Analyzes how fringe platforms produce dense ideological communities and accelerate radicalizing interaction loops.

Compares LLM-based annotation performance against expert human coders and traditional supervised models.

A large comparative study of misinformation in party communication across multiple political systems.

Examines emotional mobilization mechanisms in extremist digital milieus and their political consequences.

Theorizes how AI systems reshape state perception, categorization, and intervention in social life.

Critically reviews LLM-enabled social simulation and argues for stronger empirical validation standards.