About me
I’m a Ph.D. candidate at The University of Manchester and the NaCTeM group. I’m advised by Prof. Sophia Ananiadou. My PhD journey started in July 2023. I obtained my M.Sc. degree in Computer Science from The University of Chinese Academy of Sciences, China.
I am expected to graduate in Fall 2026 and am looking for full-time positions in AI, in either academia or industry. Please feel free to reach out if my background aligns with your needs.Research interests
My current research interests focus on Natural Language Processing and Large Language Models. Especially post-training, reinforcement learning, affective analysis, misinformation detection, and AI Safety.
In case of interest in collaboration or joint work, please feel free to reach out via zhiweiliu0810@gmail.com.
News
- Sep 2025. Invited to serve as the LREC 2026 Area Chair.
- Aug 2025. Our paper “Selective Preference Optimization via Token-Level Reward Function Estimation” has been accepted by EMNLP 2025.
- Aug 2025. Awarded a Student Grant from The Society for the Study of Artificial Intelligence and Simulation of Behaviour (AISB)
- Jul 2025. Our paper “ConspEmoLLM-v2: A robust and stable model to detect sentiment-transformed conspiracy theories” has been accepted by ECAI 2025 Download.
- May 2025. New preprint: “MMAFFBen: A Multilingual and Multimodal Affective Analysis Benchmark for Evaluating LLMs and VLMs” at arXiv.
- May 2025. Our paper “RAEmoLLM: Retrieval Augmented LLMs for Cross-Domain Misinformation Detection Using In-Context Learning based on Emotional Information” has been accepted by ACL 2025 (Main).
- Feb 2025. Invited as a speaker at ELLIS Workshop on Misinformation Detection.
Research Projects
Manchester-Melbourne-Toronto Research Project (Team Lead under Advisor) Dec 2023 - Feb 2025
Explain the strong links between emotions and misinformation and develop LLM-based methods for various misinformation detection.
Related outputs: RAEmoLLM, Rumor Detection, A Survey
Misinformation Digital Trust and Democracy II (Team Lead under Advisor) Jan 2025 - Jul 2025
Improve the robustness of LLMs for sentiment-transformed misinformation attacks.
Related outputs: ConspEmoLLM-V2
Misinformation Digital Trust and Democracy I (Team Lead under Advisor) Jan 2024 - Jul 2024
The main task is training LLMs for conspiracy theory detection using affective information
Related outputs: EmoLLMs, ConspEmoLLM
Professional services
Organizer
MisD: Workshop on Misinformation Detection in the Era of LLMs @ICWSM2025 (Leader)
Financial Misinformation Detection (FMD) challenge @COLING2025 (Leader)
Speakers
- ELLIS Workshop on Misinformation Detection
- Digital Trust and Society (CDTS) Seed Corn Showcase 2023-24
- NVIDIA Workshop
Area Chair
- LREC’26
Program Committee/Reviewer
- NeurIPS’25
- ARR’25
- SIGKDD’24, 25
- COLING’25
- WWW’25, 26
- AAAI’26
Journal Reviewer
- IEEE Transactions on Affective Computing
- Information Processing & Management
- Neurocomputing
- Applied Soft Computing Journal
- International Journal of Information Management Data Insights
- ICT Express
- HELIYON
Awards and Achievements
- Student Grant from The Society for the Study of Artificial Intelligence and Simulation of Behaviour (AISB)
- Turing Scheme Funding for Study and Work Abroad
- Scholar Award from the Department of Computer Science at the University of Manchester (072023-122026)