I am currently a Senior Agentic AI Applied Scientist at the CBU Technology Department of Taobao & Tmall Group (Alibaba) in Hangzhou, China, where I serve as Tech Lead of the Multimodal Team, driving trillion-parameter–scale RL models and architecting a multimodal agent collaboration framework for coordinated perception, reasoning, and action. In parallel, I am a researcher and PI at the Hub of Computing & Data Science, University of Hamburg, working with Prof. Chris Biemann.

I received my Ph.D. summa cum laude from the University of Hamburg in May 2026, advised by Prof. Chris Biemann. My doctoral thesis is “Bridging Vision, Language, and Gaze for Trustworthy Foundation Models”. Previously, I obtained my M.Eng. (2019) and B.Eng. (2016) degrees from the School of Computer Science and Engineering, South China University of Technology.

My research focuses on large language models and agentic systems, with an emphasis on training and alignment, evaluation and interpretability, and multilingual and multimodal learning for real-world applications. I have published 20+ papers in top international AI venues such as ACL, EMNLP, NAACL, COLING and ECAI .

💌 hanfeng.wxt@alibaba-inc.com / xintong.wang@uni-hamburg.de / m.e.xintong@gmail.com

🤝 Opening: We are hiring full-time positions & interns at Alibaba, and student assistants at the University of Hamburg. Self-motivated students with experience in LLMs, LVLMs, and Agents are welcome to reach out. Please feel free to drop me an email — I am always open to collaborations.

🔥 News

• 2026.06:  💼 I joined the CBU Technology Department, Taobao & Tmall Group (Alibaba) as a Senior Agentic AI Applied Scientist in Hangzhou.
• 2026.05:  🎓 I successfully defended my Ph.D. dissertation at the University of Hamburg (summa cum laude).
• 2025.08:  🎉 One paper accepted to EMNLP 2025 (Main, Top 15%)!
• 2025.05:  🎉 Three papers accepted to ACL 2025!
• 2024.12:  😊 Serving as an Area Chair for ACL ARR / NAACL 2025!
• 2024.11:  🎉 One paper accepted to COLING 2025 (Oral)!
• 2024.05:  🎉 One paper accepted to ACL 2024 (Findings)!
• 2024.03:  🎉 One paper accepted to LREC-COLING NeusymBridge 2024!
• 2023.08:  🛠 Co-organizer of CLEF-2024 SemEval Task: Multilingual Text Detoxification.
• 2023.07:  🎉 One paper accepted to ECAI 2023!
• 2023.06:  🎤 Invited talk at the University of Wuppertal.
• 2022.12:  ✈️ Visiting Researcher at the Institute of Psychology, Chinese Academy of Sciences.
• 2022.05:  🎉 One paper accepted to LREC 2022!
• 2021.04:  🎓 Started as a Ph.D. Candidate at the University of Hamburg.

📝 Publications

EMNLP 2025

Chinese Toxic Language Mitigation via Sentiment Polarity Consistent Rewrites

Xintong Wang, Yixiao Liu, Jingheng Pan, Liang Ding, Longyue Wang, Chris Biemann

The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025, Main, Top 15%)

• We present TOXIREWRITECN, the first Chinese detoxification dataset that explicitly preserves sentiment polarity, with 1,556 carefully annotated triplets covering five real-world scenarios.
• Comprehensive evaluation across 17 commercial and open-source LLMs reveals key limitations in sentiment-aware detoxification.

ACL 2025

CogSteer: Cognition-Inspired Selective Layer Intervention for Efficiently Steering Large Language Models

Xintong Wang, Jingheng Pan, Liang Ding, Longyue Wang, Longqin Jiang, Xingshan Li, Chris Biemann

Findings of the Association for Computational Linguistics (ACL 2025)

• We leverage eye-movement measures to analyze the layer-wise behavior of LLMs and propose a heuristic strategy for selecting the optimal steering layer for semantic intervention.
• Our framework requires only 1/N of LLM parameters and achieves +1.85% gain in toxification and +13.45% in detoxification compared to last-layer intervention.

ACL 2024

Mitigating Hallucinations in Large Vision-Language Models with Instruction Contrastive Decoding

Xintong Wang, Jingheng Pan, Liang Ding, Chris Biemann

Findings of the Association for Computational Linguistics (ACL 2024)

• A training-free Instruction Contrastive Decoding (ICD) method that contrasts standard and disturbance-instruction distributions to suppress hallucinations in LVLMs.
• Significantly mitigates both object-level and attribute-level hallucinations on POPE, MME, and LLaVA-Bench, while improving general perception and recognition.

Full Publication List

(* denotes equal contribution)

• Preprint 2025 Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models, Yunxin Li, Zhenyu Liu, Zitao Li, Xuanyu Zhang, Zhenran Xu, Xinyu Chen, Haoyuan Shi, Shenyuan Jiang, Xintong Wang, Jifang Wang, Shouzheng Huang, Xinping Zhao, Borui Jiang, Lanqing Hong, Longyue Wang, Zhuotao Tian, Baoxing Huai, Wenhan Luo, Zheng Zhang, Baotian Hu, Min Zhang. Under Review at ACM Computing Surveys.
• Preprint 2025 Rethinking Multilingual Vision-Language Translation: Dataset, Evaluation, and Adaptation, Xintong Wang, Jingheng Pan, Yixiao Liu, Xiaohu Zhao, Chenyang Lyu, Minghao Wu, Chris Biemann, Longyue Wang, Linlong Xu, Weihua Luo, Kaifu Zhang. Under Review (ACL ARR).
• Preprint 2025 A Multimodal Dataset for Visually Grounded Ambiguity in Machine Translation, Jingheng Pan, Xintong Wang, Longyue Wang, Liang Ding, Weihua Luo, Chris Biemann. Under Review (ACL ARR).
• Preprint 2025 The Bitter Lesson Learned from 2,000+ Multilingual Benchmarks, Minghao Wu, Weixuan Wang, Sinuo Liu, Huifeng Yin, Xintong Wang, Yu Zhao, Chenyang Lyu, Longyue Wang, Weihua Luo, Kaifu Zhang. Under Review (ACL ARR).
• EMNLP 2025 Chinese Toxic Language Mitigation via Sentiment Polarity Consistent Rewrites, Xintong Wang, Yixiao Liu, Jingheng Pan, Liang Ding, Longyue Wang, Chris Biemann. Main Conference, Top 15%.
• ACL 2025 CogSteer: Cognition-Inspired Selective Layer Intervention for Efficiently Steering Large Language Models, Xintong Wang, Jingheng Pan, Liang Ding, Longyue Wang, Longqin Jiang, Xingshan Li, Chris Biemann. Findings.
• ACL 2025 Metagent-P: A Neuro-Symbolic Planning Agent with Metacognition for Open Worlds, Yanfang Zhou, Yuntao Liu, Xiaodong Li, Yongqiang Zhao, Xintong Wang, Qingyu Wu, Jinlong Tian, Zhenyu Li, Xinhai Xu. Findings.
• ACL 2025 M2PA: A Multi-Memory Planning Agent for Open Worlds Inspired by Cognitive Theory, Yanfang Zhou, Xiaodong Li, Yuntao Liu, Yongqiang Zhao, Xintong Wang, Zhenyu Li, Jinlong Tian, Xinhai Xu. Findings.
• COLING 2025 Multilingual and Explainable Text Detoxification with Parallel Data, Daryna Dementieva, Daniil Moskovskiy, Nikolai Babakov, Abinew Ali Ayele, Naquee Rizwan, Florian Schneider, Xintong Wang, Seid Muhie Yimam, Dmitry Ustalov, Elisei Stakovskii, Alisa Smirnova, Ashraf Elnagar, Animesh Mukherjee, Alexander Panchenko. Oral.
• ACL 2024 Mitigating Hallucinations in Large Vision-Language Models with Instruction Contrastive Decoding, Xintong Wang, Jingheng Pan, Liang Ding, Chris Biemann. Findings.
• CLEF 2024 Overview of PAN 2024: Multi-Author Writing Style Analysis, Multilingual Text Detoxification, Oppositional Thinking Analysis, and Generative AI Authorship Verification, Abinew Ali Ayele, Nikolay Babakov, Janek Bevendorff, Xavier Bonet Casals, Berta Chulvi, Daryna Dementieva, …, Xintong Wang, Matti Wiegmann, Seid Muhie Yimam, Eva Zangerle.
• CLEF 2024 Overview of the Multilingual Text Detoxification Task at PAN 2024, Daryna Dementieva, Daniil Moskovskiy, Nikolai Babakov, Abinew Ali Ayele, Naquee Rizwan, Florian Schneider, Xintong Wang, Seid Muhie Yimam, et al.
• LREC-COLING 2024 Probing Large Language Models from a Human Behavioral Perspective, Xintong Wang, Xiaoyu Li, Xingshan Li, Chris Biemann. NeusymBridge Workshop.
• ECAI 2023 Using Self-Supervised Dual Constraint Contrastive Learning for Cross-modal Retrieval, Xintong Wang, Xiaoyu Li, Liang Ding, Sanyuan Zhao, Chris Biemann.
• AACL-IJCNLP SRW 2022 Language over Labels: Contrastive Language Supervision Exceeds Purely Label-Supervised Classification Performance on Chest X-Rays, Anton Orell Wiehe, Florian Schneider, Sebastian Blank, Xintong Wang, Hans-Peter Zorn, Chris Biemann.
• LREC 2022 MOTIF: Contextualized Images for Complex Words to Improve Human Reading, Xintong Wang*, Florian Schneider*, Özge Alaçam, Prateek Chaudhury, Chris Biemann. [Dataset]
• NAACL SRW 2021 Towards Multi-Modal Text-Image Retrieval to Improve Human Reading, Florian Schneider, Özge Alaçam, Xintong Wang, Chris Biemann.
• Information Sciences 2020 Plausibility-promoting Generative Adversarial Network for Abstractive Text Summarization with Multi-task Constraint, Min Yang, Xintong Wang, Yao Lu, Jianming Lv, Ying Shen, Chengming Li. JCR-Q1.

🧑‍🏫 Teaching

• Winter 2025 & 2026, Lecturer, Exercises Natural Language Processing and the Web (Master), University of Hamburg.
• Summer 2025 & 2026, Lecturer, Exercises Statistical Methods of Language Technology (Master), University of Hamburg.
• Winter 2024, Teaching Assistant, Introduction to Python for Research (Bachelor / Master / PhD), Max Planck Institute.
• Winter 2024, Co-Instructor, Deep Learning for Natural Language Processing (Bachelor), University of Hamburg.
• Summer 2022 & 2023, Project Mentor, Web Interfaces for Language Processing Systems (Master Project), University of Hamburg.
• Winter 2020, Co-Lecturer, Natural Language Processing and the Web (Master), University of Hamburg.

🎓 Supervision

Co-Supervised Doctoral Students

• Jingheng Pan (Ph.D. Candidate, Universität Hamburg, 2025 – present) — LVLM Post-Training, Reasoning.
• Yixiao Liu (Ph.D. Candidate, Universität Hamburg, 2026 – present) — LLM Post-Training, Interpretability, Cognitive Analysis.

Supervised Master Students

• Longqin Jiang (M.Sc., Universität Hamburg, 2025 – present) — In-Image Text Understanding, Vision-Language Models.
• Jingfan Xin (M.Sc., Universität Hamburg, 2025 – present) — Large Reasoning Models.

Interns

• Duo Li (Ph.D. Candidate, Nanyang Technological University, 2026 Summer) — Multimodal Agents.
• Xincheng Li (Ph.D. Candidate, Tongji University, 2026 Summer) — LLM Post-Training, RL.
• Haonan Qi (M.Sc., Shanghai Jiao Tong University, 2026 Summer) — Vision-Language Reasoning.
• Yongqi Zhang (M.Sc., Southeast University, 2026 Summer) — Multimodal LLMs.

Previous Students

• Xiaoyu Li (M.Sc., TU Berlin & Beijing Institute of Technology, 2024) — Foundation Models, Cross-Modal Representation Learning.
• Fabian Meyer (M.Sc., Universität Hamburg, 2023) — Out-of-Distribution Detection, Robustness.
• Anton Orell Wiehe (M.Sc., Universität Hamburg, 2022) — Domain Adaptation, Multi-Modal Foundation Models.
• Matthew Ng Cher-Wai (M.Sc., Universität Hamburg, 2022) — Multi-Modal Generation, Transformers.
• Ankit Srivastava (M.Sc., Universität Hamburg, 2022) — Lexical Simplification, Educational NLP.
• Florian Schneider (M.Sc., Universität Hamburg, 2021) — Self-Supervised Multi-Modal Retrieval. GSCL Best Master’s Thesis Award 2023.
• Prateek Chaudhury (Research Assistant, IIT Delhi, 2021) — Educational NLP, Reading Comprehension.

🎖 Honors and Awards

• 2025, Alibaba Group AliStar Program — Offer declined.
• 2025, Alibaba International AliStar Program — Offer declined.
• 2025, Huawei Genius Youth Program — Offer declined.
• 2025, Ant Group AntStar Plan A Program — Offer declined.
• 2023, ECAI 2023 Travel Grant Award, 26th European Conference on Artificial Intelligence.

💬 Invited Talks

• 2025.03, Multimodal Representation Learning: Understanding and Leveraging. Host: Prof. Jianming Lv, South China University of Technology.
• 2024.08, Towards Truthfulness, Safeness, and Explainable Advanced Foundation Models. Host: Prof. Xingshan Li, Institute of Psychology, Chinese Academy of Sciences.
• 2024.07, Mitigating Hallucinations in Large Vision-Language Models (LVLMs). Online Talk.
• 2024.05, Probing Large Language Models from a Human Behavioral Perspective. Host: Dr. Tiansi Dong, NeusymBridge @ LREC-COLING 2024.
• 2023.06, Probing Large Language Models (LLMs) for Predicting Human Behavioral Data. Hosts: Prof. Markus Hofmann & Prof. Ralph Radach, Universität Wuppertal.

🛠 Academic Services

Funding Reviewer

• Vienna Science and Technology Fund — Vienna Research Groups for Young Investigators (2025).

Workshop & Shared Task Organization

• POLAR @ SemEval 2026 — Attitude Polarization Detection in Multilingual Text.
• CLEF 2025 SemEval Task — Multilingual Text Detoxification.
• CLEF 2024 SemEval Task — Multilingual Text Detoxification.

Area Chair

• ACL Rolling Review (ARR) — Multimodal Learning, 2024 – 2025.

Session Chair

• ECAI 2023 — Speech and Natural Language Processing.

Conference Reviewer

• ACL, EMNLP, NAACL, EACL, COLING, LREC-COLING, IJCNLP-AACL, AAAI, IJCAI, NeurIPS, AISTATS, CVPR, ECCV, ACM MM, ECAI — 2019 – present.
• IJCAI 2021 — Senior Program Committee.

Journal Reviewer

• ACM TALLIP (2020 – 2021); IEEE Access (2020 – 2021).

🐈 Misc.

I have a cat called Minus — his name is the German word for subtraction, with my hope that he can live life on easy mode.