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

arXiv: https://arxiv.org/abs/2605.02712 · HTML: https://mlsanity.com/paper/2605.02712/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 3 of its 12 references that are in the map.
- First posted on arXiv 4 May 2026.

## Builds on (most cited first, 3 of 3)

- [IMGTB: A Framework for Machine-Generated Text Detection Benchmarking](https://mlsanity.com/paper/2311.12574/) (2023)
- [Increasing the Robustness of the Fine-tuned Multilingual Machine-Generated Text Detectors](https://mlsanity.com/paper/2503.15128/) (2025)
- [mdok of KInIT: Robustly Fine-tuned LLM for Binary and Multiclass AI-Generated Text Detection](https://mlsanity.com/paper/2506.01702/) (2025)

## Abstract (arXiv)

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.