Spaces:
Running
Running
| title: ProBERT Demo | |
| emoji: π― | |
| colorFrom: blue | |
| colorTo: green | |
| sdk: gradio | |
| sdk_version: 6.5.1 | |
| app_file: app.py | |
| pinned: false | |
| short_description: Detect rhetorical overconfidence in text | |
| license: other | |
| license_name: collapse-index-open-model-license | |
| license_link: https://huggingface.co/collapseindex/ProBERT-1.0/blob/main/LICENSE.md | |
| # ProBERT v1.0 - Rhetorical Confidence Detection | |
| **Demo for [collapseindex/ProBERT-1.0](https://huggingface.co/collapseindex/ProBERT-1.0)** | |
| ProBERT detects rhetorical overconfidence patterns in text, classifying across three categories: | |
| - π’ **process_clarity** - Clear procedural logic with explicit steps | |
| - π **rhetorical_confidence** - Assertive claims without supporting evidence | |
| - π΄ **scope_blur** - Vague generalizations or abstract statements | |
| ## Performance | |
| - **Accuracy:** 95.6% | |
| - **Inference Speed:** 31ms (CPU) | |
| - **Type I Errors (Collapse Index):** 0 (behaviorally stable under perturbations) | |
| ## License | |
| [Collapse Index Open Model License v1.0](https://huggingface.co/collapseindex/ProBERT-1.0/blob/main/LICENSE.md) (Permissive use + Attribution) | |
| ## Contact | |
| - Website: https://collapseindex.org | |
| - GitHub: https://github.com/collapseindex/ProBERT-1.0 | |
| - Email: ask@collapseindex.org | |