AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

TSM-Bench: Detecting LLM-Generated Text in Real-World Wikipedia Editing Practices

Gerrit Quaremba, Elizabeth Black, Denny Vrandecic, Elena Simperl

ICLR 2026 · Large language models › Language, safety and interpretability · approach: Zero-shot machine-text detection · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
evaluate on task-specific MGT instead of generic MGToff-the-shelf detectors · Wikipedia MGT (GPT-4o, English)accuracy: >93% -> between 47% and 73%Hurt
use supervised detectors instead of zero-shotdetector family · Task-specific MGTaccuracy: zero-shot max 64.7% vs supervised 79.7% to 91.8%Improved
go from sentence-level to paragraph-level textsupervised detectors, TST · Text Style Transfer (English)accuracy: +15.7% for supervised detectorsImproved
use paragraph-continuation (human-machine mixed) text instead of introductory paragraphszero-shot detectors · Paragraph Continuationaccuracy: Binoculars 57.8 -> 52.1 (English)Hurt
move language from English to VietnameseBinoculars · Paragraph Continuationaccuracy: 52.1 -> 60Improved
train on task-specific data rather than generic datamDeBERTa · Cross-domain test setsaccuracy: 73.3 -> 89.7Improved
train on generic data and test on task-specific datamDeBERTa · TSM-Bench (English)accuracy: 92.8-100 diagonal -> at most 73.3Hurt
train on one task and test on othersmDeBERTa · Cross-task (English)accuracy: 72.1 (Summarisation) vs 60.5 (Introductory Paragraph), near chance otherwiseHurt
add retrieval context (RAG) to paragraph promptGPT-4o mini · Introductory Paragraph prompt eval (English)RougeL: 0.47 (+0.29)Improved
add one-shot examples to summarisation promptGPT-4o mini · Summarisation prompt eval (English)RougeL: 0.36 (+0.10)Improved
add five-shot examples to TST promptGPT-4o mini · TST prompt eval (English)BLEU: 0.55 (+0.21)Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy