mdok-style at SemEval-2026 Task 10: Finetuning LLMs for Conspiracy Detection

SemEval-2026 Task 10 is focused on conspiracy detection. Specifically, the goal is to detect whether a Reddit comment expresses a conspiracy belief. Our submitted mdok-style systemutilizes data augmentation and self-training (to cope with a rather small amount of training data) to finetune the Qwen3-32B model for a binary text-classification task. The submitted system is very competitive, ranking in the 85th percentile (8th out of 52 submissions). The results shown that our approach, which originated in machine-generated text detection, can be used for conspiracy detection as well.

Cite: Dominik Macko. 2026. mdok-style at SemEval-2026 Task 10: Finetuning LLMs for Conspiracy Detection. In Proceedings of the 20th International Workshop on Semantic Evaluation (2026), pages 446–450, San Diego, California, USA. Association for Computational Linguistics.