English

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

Computation and Language 2026-05-05 v1 Artificial Intelligence

Abstract

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 system utilizes 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.

Keywords

Cite

@article{arxiv.2605.02712,
  title  = {mdok-style at SemEval-2026 Task 10: Finetuning LLMs for Conspiracy Detection},
  author = {Dominik Macko},
  journal= {arXiv preprint arXiv:2605.02712},
  year   = {2026}
}

Comments

Accepted to the SemEval-2026 workshop of the ACL 2026 conference