Qwen it detect machine-generated text?

TL;DR

This paper presents a multilingual approach for detecting machine-generated text, achieving top results in the Coling 2025 GenAI Workshop with models based on masked and causal language modeling.

Contribution

The team developed and evaluated models for multilingual machine-generated text detection, achieving state-of-the-art performance in a competitive benchmark.

Findings

Achieved first place in F1 Micro score

Secured second place in F1 Macro score

Demonstrated effectiveness of masked and causal models

Abstract

This paper describes the approach of the Unibuc - NLP team in tackling the Coling 2025 GenAI Workshop, Task 1: Binary Multilingual Machine-Generated Text Detection. We explored both masked language models and causal models. For Subtask A, our best model achieved first-place out of 36 teams when looking at F1 Micro (Auxiliary Score) of 0.8333, and second-place when looking at F1 Macro (Main Score) of 0.8301