| language: en | |
| license: mit | |
| inference: false | |
| tags: | |
| - ai-text-detection | |
| base_model: jhu-clsp/ettin-encoder-1b | |
| # MELD (mirror) | |
| This is an unmodified mirror of [`anon-review-meld-2026/meld`](https://huggingface.co/anon-review-meld-2026/meld) | |
| at revision `8990324abd92e1fa17072f6887ea1e5c1cef5abc`, kept so that | |
| [Definitely Human™](https://github.com/Ogdeveloperc/definitely-human) installs keep working if the original | |
| anonymous-review repository moves. All credit goes to the authors. | |
| > **MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text** | |
| > Chenjun Li, Cheng Wan, Haomiao Chen, Johannes C. Paetzold (Cornell University). | |
| > [arXiv:2605.06903](https://arxiv.org/abs/2605.06903) | |
| An AI-generated text detector for English (Ettin-1B / ModernBERT encoder with a custom scoring head). | |
| Usage is unchanged from the original: | |
| ```bash | |
| pip install torch transformers safetensors | |
| python meld.py "Paste the text to check here." | |
| ``` | |
| The scoring head is custom, so `pipeline()` and `AutoModel` do not apply; use `meld.py` or its `Scorer` class. | |
| Inputs should be at least about 100 words. License: MIT, as released by the authors. | |