Add model card for REVEAL

This PR adds a model card for the REVEAL model presented in the paper "Reasoning-Aware AIGC Detection via Alignment and Reinforcement".

The model card includes:
- Links to the paper, project page, and code repository.
- Metadata for the `text-classification` pipeline and the `transformers` library.

README.md

@@ -0,0 +1,19 @@
+---
+library_name: transformers
+pipeline_tag: text-classification
+---
+
+# REVEAL: Reasoning-Aware AIGC Detection via Alignment and Reinforcement
+
+REVEAL is a detection framework that generates interpretable reasoning chains before classifying text as AI-generated or human-written. It is designed to be robust against evolving Large Language Models (LLMs) and provides transparent, interpretable results.
+
+- **Paper:** [Reasoning-Aware AIGC Detection via Alignment and Reinforcement](https://huggingface.co/papers/2604.19172)
+- **Project Page:** [https://aka.ms/reveal](https://aka.ms/reveal)
+- **Repository:** [https://github.com/microsoft/AnthropomorphicIntelligence](https://github.com/microsoft/AnthropomorphicIntelligence)
+
+## Overview
+REVEAL addresses the challenge of reliable AI-generated content (AIGC) detection by introducing a reasoning-aware approach. The model uses a two-stage training strategy: 
+1. **Supervised Fine-Tuning (SFT):** To establish the model's capability to generate logical reasoning chains.
+2. **Reinforcement Learning (RL):** To improve classification accuracy, logical consistency, and reduce hallucinations.
+
+The framework was trained on the **AIGC-text-bank**, a comprehensive multi-domain dataset featuring diverse LLM sources and authorship scenarios.