Qiang Sheng盛强
Associate Professor / Researcher
Institute of Computing Technology, Chinese Academy of Sciences
Calls
About
I am an Associate Professor/Researcher at the Media Synthesis and Forensics Lab, Institute of Computing Technology, Chinese Academy of Sciences. I am also a supervisor at School of Computer Science and Technology, University of Chinese Academy of Sciences (UCAS), where I got my Ph.D. degree under the supervision of Professor Juan Cao. My research interests include:
- Application: Fake News/Misinformation Detection, Fact-Checking, Machine-Generated Content Detection
- Direction: Natural Language Understanding and Social Media Mining
- Vision: Make the World More Credible
News
View All →Our paper on fine-grained LLM-generated text detection received ACL 2026 Outstanding Paper Award and is selected as an SAC Highlight.
One paper got accepted by MM 2026.
One co-authored paper got accepted by KDD 2026.
One co-authored paper got accepted by ICML 2026.
Two co-authored papers got accepted by ACL 2026.
Selected Publications
View All →MM'26Multi-Tool Image Editing Attribution in Facial Forgery
Sheng Liu, Qiang Sheng, Danding Wang, Yu Li, Chenming Zhou, Juan Cao
Proceedings of the 34th ACM International Conference on Multimedia
We build the benchmark MultiEdit and propose a method to judge whether a tool participated in the mutli-turn editing process of a manipulated facial image.
KDD'26EvoFEND: Dual Memory-Driven Self-Evolving Fake News Detection
Beizhe Hu, Qiang Sheng, Hao Mi, Jiaying Wu, Zhengjia Wang, Yuanlong Yu, Danding Wang, Xuming Hu, Juan Cao
Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2 (Acceptance Rate: 17.7%)
We build an agentic framework for fake news detection that can evolve itself by reading social media streams.
ICML'26FactGuard: Agentic Video Misinformation Detection via Reinforcement Learning
Zehao Li, Hongwei Yu, Hao Jiang, Qiang Sheng, Yilong Xu, Baolong Bi, Yang Li, Zhenlong Yuan, Yujun Cai, Zhaoqi Wang
Proceedings of the 43rd International Conference on Machine Learning (Acceptance Rate: 6352/23918=26.6%)
We propose an agentic misinformation video detection framework that can reason the veracity iteratively with self-refinement.
ACL'26Logical Consistency as a Bridge: Improving LLM Hallucination Detection via Label Constraint Modeling between Responses and Self-Judgments
Hao Mi, Qiang Sheng, Shaofei Wang, Beizhe Hu, Yifan Sun, Zhengjia Wang, Hengqi Zeng, Yang Li, Danding Wang, Juan Cao
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Acceptance Rate: 18.9%)
We model the logical label constraint between LLM responses and self-judgments as a bridge to enhance hallucination detection.
ACL'26Beyond the Final Actor: Modeling the Dual Roles of Creator and Editor for Fine-Grained LLM-Generated Text Detection
Yang Li, Qiang Sheng, Zhengjia Wang, Yehan Yang, Danding Wang, Juan Cao
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Acceptance Rate: 18.9%)
We disentangle the dual roles of creator and editor in LLM text generation, enabling fine-grained detection of AI-generated content at different revision modes.
AAAI'26Reasoning About the Unsaid: Misinformation Detection with Omission-Aware Graph Inference
Zhengjia Wang, Danding Wang, Qiang Sheng, Jiaying Wu, Juan Cao
Proceedings of the 40th AAAI Conference on Artificial Intelligence (Acceptance Rate: 4167/23680=17.60%)
We consider the omitted information to better reason the creator's intent for misinformation detection.
NeurIPS'25From Judgment to Interference: Early Stopping LLM Harmful Outputs via Streaming Content Monitoring
Yang Li, Qiang Sheng, Yehan Yang, Xueyao Zhang, Juan Cao
Proceedings of the 39th Annual Conference on Neural Information Processing Systems (Acceptance Rate: 5290/21575=24.52%)
We build a content moderator that can early stop LLMs' harmful outputs with low latency.
SIGIR'25LLM-Generated Fake News Induces Truth Decay in News Ecosystem: A Case Study on Neural News Recommendation
Beizhe Hu, Qiang Sheng, Juan Cao, Yang Li, Danding Wang
Proceedings of The 48th International ACM SIGIR Conference on Research and Development in Information Retrieval (Acceptance Rate: 239/1071=22.3%)
We reveal the truth-decay phenomenon where real news gradually loses its top-ranked advantage against fake news when LLM-generated news penetrates.
MM'24FakingRecipe: Detecting Fake News on Short Video Platforms from the Perspective of Creative Process
Yuyan Bu, Qiang Sheng, Juan Cao, Peng Qi, Danding Wang, Jintao Li
Proceedings of the 32nd ACM International Conference on Multimedia (Acceptance Rate: 1149/4385=26.2%)
We detect fake news on short video platforms by modeling videos from the faking process perspective and constructed a FakeSV's sister dataset in English, namely FakeTT.
AAAI'24Bad Actor, Good Advisor: Exploring the Role of Large Language Models in Fake News Detection
Beizhe Hu, Qiang Sheng, Juan Cao, Yuhui Shi, Yang Li, Danding Wang, Peng Qi
Proceedings of the 38th AAAI Conference on Artificial Intelligence
Large LMs generally underperform fine-tuned Small LMs for fake news detection, but they can be good advisors by providing rationales.