AI Detection

A french pretrained classification model that aims to detect whether an input has been generated by a human or by a Large Language Model such as Chat-GPT or LLaMA.
This model is aimed to one day help educators verify if work is done by humans or not. The model will, by design, classify as human-made text that has been generated by AI but rewritten by a human. This is because its aim is to catch pure ai without any human input.

This model is based on on BARThez.
It is pretrained on a corpus of 15 000 of french articles that have been summarized by both humans and AI

Downloads last month
163
Safetensors
Model size
140M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model authors have turned it off explicitly.

Space using Anvil-ML/detecteur-ia 1