🧪 Test?View on arXiv
Same Origin, Different Verdicts: Text Production Conditions Confound Machine-Generated Text Detection in Russian
text detectionmachine-generated textnatural language processing
2611.05385
Builder Relevance
2h ago70%
Abstract
This study investigates how text production conditions affect the detection of machine-generated text in Russian, revealing significant discrepancies in detection outcomes based on task wording.
Reality Card
Core Claim
The study demonstrates that changing task modes can lead to conflicting detection results for machine-generated text, highlighting the influence of production conditions on detection accuracy.
Method / Result
The baseline on the source identifier alone achieved a median AUROC of 1.0000 on thirteen group splits.
Limitations
The results are influenced by specific measurement systems and corpus characteristics, which may limit reproducibility across different datasets.
Paper to code
Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.
No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.