I am Heming Xia, a Ph.D. student in the NLP Group at The Hong Kong Polytechnic University, supervised by Prof. Wenjie Li. I earned my master’s degree from the MOE Key Lab of Computational Linguistics at Peking University, advised by Prof. Zhifang Sui. Prior to that, I completed my bachelor’s degree from the School of Physics at Peking University in 2020. I have also worked as a Research Intern at the NLC Group @ Microsoft Research Asia, where I had the privilege of collaborating with Dr. Tao Ge. Currently, I am an intern at SEA AI Lab, working closely with Dr. Cunxiao Du. For more details, please refer to my CV.
Research
I am broadly interested in natural language processing and machine learning. My current research focuses on 1) efficient and effective NLP, 2) tool learning, and 3) cross vision and language understanding.
News
[2025.08] Got three papers accepted by EMNLP 2025, congrats to all co-authors🎉!
[2025.05] Got three papers accepted by ACL 2025.
[2025.02] Released TokenSkip, enabling LLMs to skip less important tokens during reasoning.
[2025.01] Got one paper accepted by ICLR 2025.
[2025.01] Organized a tutorial on Speculative Decoding at COLING 2025.
[2024.10] Released SWIFT: on-the-fly self-speculative decoding for LLM inference acceleration.
[2024.09] Got four papers accepted by EMNLP 2024.
[2024.05] Got two papers accepted by ACL 2024.
[2024.01] Released Spec-Bench: a comprehensive benchmark for Speculative Decoding.
[2024.01] Released our new survey 📖 on Speculative Decoding.
[2024.01] Started my Ph.D. study at the NLP Group @ PolyU, supervised by Prof. Wenjie Li.
[2023.10] Got three papers accepted by EMNLP 2023.
Publications
Most recent publications on Google Scholar.
* indicates equal contribution
Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning
Xinghao Chen*, Anhao Zhao*, Heming Xia, Xuan Lu, Hanlin Wang, Yanjun Chen, Wei Zhang, Jian Wang, Wenjie Li, Xiaoyu Shen
Arxiv 2025. [link] [code]
KNN-SSD: Enabling Dynamic Self-Speculative Decoding via Nearest Neighbor Layer Set Optimization
Mingbo Song, Heming Xia, Jun Zhang, Chak Tou Leong, Qiancheng Xu, Wenjie Li, Sujian Li
Arxiv 2025. [link]
From Query to Logic: Ontology-Driven Multi-Hop Reasoning in LLMs
Haonan Bian, Yutao Qi, Rui Yang, Yuanxi Che, Jiaqian Wang, Heming Xia, Ranran Zhen
Arxiv 2025. [link]
TokenSkip: Controllable Chain-of-Thought Compression in LLMs
Heming Xia, Yongqi Li, Chak Tou Leong, Wenjie Wang, Wenjie Li
EMNLP 2025. [link] [code] (Best Oral Presentation @ PolyU COMP 2025 Research Conference)
SpecVLM: Enhancing Speculative Decoding of Video LLMs via Verifier-Guided Token Pruning
Yicheng Ji*, Jun Zhang*, Heming Xia, Jinpeng Chen, Lidan Shou, Gang Chen, Huan Li
EMNLP 2025. [link]
Beyond Single Frames: Can LMMs Comprehend Temporal and Contextual Narratives in Image Sequences?
Xiaochen Wang*, Heming Xia*, Jialin Song, Longyu Guan, Yixin Yang, Qingxiu Dong, Weiyao Luo, Yifan Pu, Yiru Wang, Xiangdi Meng, Wenjie Li, Zhifang Sui
Findings of EMNLP 2025. [link]
Towards Harmonized Uncertainty Estimation for Large Language Models
Rui Li, Jing Long, Muge Qi, Heming Xia, Lei Sha, Peiyi Wang, Zhifang Sui
ACL 2025. [link] (Oral Presentation)
How Far are LLMs from Being Our Digital Twins? A Benchmark for Persona-Based Behavior Chain Simulation
Rui Li, Heming Xia, Xinfeng Yuan, Qingxiu Dong, Lei Sha, Wenjie Li, Zhifang Sui
Findings of ACL 2025. [link] [code]
PEToolLLM: Towards Personalized Tool Learning in Large Language Models
Qiancheng Xu, Yongqi Li, Heming Xia, Fan Liu, Min Yang, Wenjie Li
Findings of ACL 2025. [link] [code]
SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference Acceleration
Heming Xia, Yongqi Li, Jun Zhang, Cunxiao Du, Wenjie Li
ICLR 2025. [link] [code]
AppBench: Planning of Multiple APIs from Various APPs for Complex User Instruction
Hongru Wang, Rui Wang, Boyang Xue, Heming Xia, Jingtao Cao, Zeming Liu, Jeff Z. Pan, Kam-Fai Wong
EMNLP 2024. [link] [code]
A Survey on In-context Learning
Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Jingyuan Ma, Rui Li, Heming Xia, Jingjing Xu, Zhiyong Wu, Baobao Chang, Xu Sun, Lei Li, Zhifang Sui
EMNLP 2024. [link] [code] [机器之心]
Enhancing Tool Retrieval with Iterative Feedback from Large Language Models
Qiancheng Xu, Yongqi Li, Heming Xia, Wenjie Li
Findings of EMNLP 2024. [link]
Taking a Deep Breath: Enhancing Language Modeling of Large Language Models with Sentinel Tokens
Weiyao Luo, Suncong Zheng, Heming Xia, Weikang Wang, Yan Lei, Tianyu Liu, Shuang Chen, Zhifang Sui
Findings of EMNLP 2024. [link]
Unlocking Efficiency in Large Language Model Inference: A Comprehensive Survey of Speculative Decoding
Heming Xia, Zhe Yang, Qingxiu Dong, Peiyi Wang, Yongqi Li, Tao Ge, Tianyu Liu, Wenjie Li, Zhifang Sui
Findings of ACL 2024. [link] [code] [机器之心]
Can Large Multimodal Models Uncover Deep Semantics Behind Images?
Yixin Yang, Zheng Li, Qingxiu Dong, Heming Xia, Zhifang Sui
Findings of ACL 2024. [link] [code]
ImageNetVC: Zero- and Few-Shot Visual Commonsense Evaluation on 1000 ImageNet Categories
Heming Xia*, Qingxiu Dong*, Lei Li, Jingjing Xu, Tianyu Liu, Ziwei Qin, Zhifang Sui
Findings of EMNLP 2023. [link] [code]
Bi-Drop: Enhancing Fine-tuning Generalization via Synchronous sub-net Estimation and Optimization
Shoujie Tong*, Heming Xia*, Damai Dai, Runxin Xu, Tianyu Liu, Binghuai Lin, Yunbo Cao, and Zhifang Sui
Findings of EMNLP 2023. [link]
Speculative Decoding: Exploiting Speculative Execution for Accelerating Seq2seq Generation
Heming Xia*, Tao Ge*, Peiyi Wang, Si-Qing Chen, Furu wei, and Zhifang Sui
Findings of EMNLP 2023. [link] [code]
Enhancing Continual Relation Extraction via Classifier Decomposition
Heming Xia, Peiyi Wang, Tianyu Liu, Binghuai Lin, Yunbo Cao, and Zhifang Sui
Findings of ACL 2023. [link] [code]
Premise-based Multimodal Reasoning: Conditional Inference on Joint Textual and Visual Clues
Qingxiu Dong*, Ziwei Qin*, Heming Xia, Tian Feng, Shoujie Tong, Haoran Meng, Lin Xu, Zhongyu Wei, Weidong Zhan, Baobao Chang, Sujian Li, Tianyu Liu and Zhifang Sui
ACL 2022. [link] [code]
Improved deep learning techniques in gravitational-wave data analysis
Heming Xia, Lijing Shao, Junjie Zhao and Zhoujian Cao
Phys. Rev. D 103 2021. [link] [code]
Service
Area Chair/Action Editor:
2025: ACL, EMNLP, ACL ARR
Reviewer/Program Committee Member:
2025: ICLR, ICML, NeurIPS, ACM MM, TASLP, TWEB
2024: ACL, EMNLP (Outstanding Reviewer🌟), NAACL, ACL ARR
2023: AACL, NAACL, ACL ARR
2022: NeurIPS, AACL
Teaching Assistant:
COMP 5423: Natural Language Processing, Fall 2025, PolyU
COMP 5423: Natural Language Processing, Spring 2025, PolyU
COMP 5140: Metaverse Fundamentals, Fall 2024, PolyU
COMP 2S01: Technology Beyond Borders: Service Learning Across Cultural and Ethnic, Spring 2024, PolyU
Invited Talks
[2025.06] Sharing Panel: Efficient Reasoning in Large Language Models at NICE and MLNLP. [video]
[2025.05] Stop Overthinking: Towards Efficient Reasoning in Large Language Models at Theory Lab, Huawei Hong Kong Research Center. [slides]
[2025.01] Speculative Decoding for Efficient LLM Inference at COLING 2025. [homepage] [slides] [video]
[2024.03] Unlocking the Efficiency of LLM Inference: A Comprehensive Survey of Speculative Decoding at NICE and CIP Group @CASIA. [video] [slides]
Awards
Merit Student, Peking University (2021)
Scholarship of National Astronomical Observatory, Chinese Academy of Sciences (2019)
Merit Student, Henan Province, China (2016)