Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Detecting Subtle Differences between Human and Model Languages Using Spectrum of Relative Likelihood

About

Human and model-generated texts can be distinguished by examining the magnitude of likelihood in language. However, it is becoming increasingly difficult as language model's capabilities of generating human-like texts keep evolving. This study provides a new perspective by using the relative likelihood values instead of absolute ones, and extracting useful features from the spectrum-view of likelihood for the human-model text detection task. We propose a detection procedure with two classification methods, supervised and heuristic-based, respectively, which results in competitive performances with previous zero-shot detection methods and a new state-of-the-art on short-text detection. Our method can also reveal subtle differences between human and model languages, which find theoretical roots in psycholinguistics studies. Our code is available at https://github.com/CLCS-SUSTech/FourierGPT

Yang Xu, Yu Wang, Hao An, Zhichen Liu, Yongyuan Li• 2024

Related benchmarks

TaskDatasetResultRank
LLM-generated text detectionEvoBench
LLaMA3 Score63.99
26
Machine-generated text detectionMAGE--
18
Machine-generated text detectionDetectRL Training Text: ChatGPT--
12
AI-generated text detectionReuters
GPT4All Score99.28
8
AI-generated text detectionEssay
AUROC (GPT4All)91.7
8
Machine-generated text detectionDetectRL Llama-2-70b
AUROC0.5811
6
Machine-generated text detectionDetectRL Google-PaLM
AUROC59.99
6
Showing 7 of 7 rows

Other info

Follow for update