Saved in:
Bibliographic Details
Main Authors: Wang, Zhao, Xiong, Max, Lian, Jianxun, Dou, Zhicheng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.19172
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917425355685888
author Wang, Zhao
Xiong, Max
Lian, Jianxun
Dou, Zhicheng
author_facet Wang, Zhao
Xiong, Max
Lian, Jianxun
Dou, Zhicheng
contents The rapid advancement and widespread adoption of Large Language Models (LLMs) have elevated the need for reliable AI-generated content (AIGC) detection, which remains challenging as models evolve. We introduce AIGC-text-bank, a comprehensive multi-domain dataset with diverse LLM sources and authorship scenarios, and propose REVEAL, a detection framework that generates interpretable reasoning chains before classification. Our approach uses a two-stage training strategy: supervised fine-tuning to establish reasoning capabilities, followed by reinforcement learning to improve accuracy, improve logical consistency, and reduce hallucinations. Extensive experiments show that REVEAL achieves state-of-the-art performance across multiple benchmarks, offering a robust and transparent solution for AIGC detection. The project is open-source at https://aka.ms/reveal
format Preprint
id arxiv_https___arxiv_org_abs_2604_19172
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Reasoning-Aware AIGC Detection via Alignment and Reinforcement
Wang, Zhao
Xiong, Max
Lian, Jianxun
Dou, Zhicheng
Artificial Intelligence
The rapid advancement and widespread adoption of Large Language Models (LLMs) have elevated the need for reliable AI-generated content (AIGC) detection, which remains challenging as models evolve. We introduce AIGC-text-bank, a comprehensive multi-domain dataset with diverse LLM sources and authorship scenarios, and propose REVEAL, a detection framework that generates interpretable reasoning chains before classification. Our approach uses a two-stage training strategy: supervised fine-tuning to establish reasoning capabilities, followed by reinforcement learning to improve accuracy, improve logical consistency, and reduce hallucinations. Extensive experiments show that REVEAL achieves state-of-the-art performance across multiple benchmarks, offering a robust and transparent solution for AIGC detection. The project is open-source at https://aka.ms/reveal
title Reasoning-Aware AIGC Detection via Alignment and Reinforcement
topic Artificial Intelligence
url https://arxiv.org/abs/2604.19172