Wenqi Pei
Portrait
HKU ECE MPhil

I am a MPhil student at The University of Hong Kong (HKU), supervised by Prof. Hongyang Du at NICE Lab. Prior to this, I completed my Bachelor's degree at the National University of Singapore (NUS) under the supervision of Prof. Bingsheng He. My research focuses on various aspects of Agentic AI, with a particular emphasis on Data Agent. I have also begun exploring multimodal Video-LLMs. I am open to collaborating with people who share similar research interests.

I am currently seeking PhD opportunities for Fall 2027. Please feel free to reach out if you see a potential research fit.

Google Scholar GitHub Twitter
Education
  • The University of Hong Kong
    The University of Hong Kong
    Master of Philosophy
    Sep. 2025 - Present
  • National University of Singapore
    National University of Singapore
    Bachelor of Computing (Honours), Minor: Economics
    Aug. 2021 - Jul. 2025
Honors & Awards
  • Dean's List
    Sem2 2024-2025
  • Top Student for CS Research Methodology
    Sem1 2024-2025
News
2026
TempCloze accepted to EMNLP Finding 🎉
Aug 23
Released our survey on efficient tool-using LLM agents 📚
Aug 06
ROSE accepted to ACL Main 🎉
Apr 04
2025
NL2SQLBench accepted to VLDB 🎉
Dec 16
Feather-SQL accepted to IJCNLP-AACL Finding 🎉
Oct 25
Started graduate study at HKU 🎓
Sep 01
InterFeedback accepted to EMNLP Finding 🎉
Aug 20
Self-Monitoring Large Language Models accepted to ACM TOIS 🎉
Aug 17
Visited HKUST(GZ) DIAL Lab directed by Prof. Yuyu Luo for one month 🔬
Aug 01
Graduated from NUS with BComp (Honours) 🎓
Jul 15
Selected Publications (view all )
TempCloze: Can Video-LLMs Identify the Missing Middle?
TempCloze: Can Video-LLMs Identify the Missing Middle?

Wenqi Pei*, Hengyuan Zhao*, Yilai Liu*, Jiahao Meng, Han Chen, Ziyu Wang, Hongyang Du (* equal contribution)

EMNLP Finding 2026

We introduce TempCloze, a video cloze benchmark for evaluating visual temporal reasoning in Video-LLMs by asking them to identify the true missing middle between beginning and ending clips. Across 1,521 videos and 31 Video-LLMs, TempCloze reveals temporal alignment as the primary bottleneck.

[Paper] [Code] [Dataset]

TempCloze: Can Video-LLMs Identify the Missing Middle?

Wenqi Pei*, Hengyuan Zhao*, Yilai Liu*, Jiahao Meng, Han Chen, Ziyu Wang, Hongyang Du (* equal contribution)

EMNLP Finding 2026

We introduce TempCloze, a video cloze benchmark for evaluating visual temporal reasoning in Video-LLMs by asking them to identify the true missing middle between beginning and ending clips. Across 1,521 videos and 31 Video-LLMs, TempCloze reveals temporal alignment as the primary bottleneck.

[Paper] [Code] [Dataset]

How to Make Tool-Using LLM Agents Efficient? A Survey
How to Make Tool-Using LLM Agents Efficient? A Survey

Yunuo Hu*, Ziyang Cheng*, Wenqi Pei*‡, Miaomiao Li, Boyan Li, Yizhang Zhu, Yupeng Xie, Han Chen, Shizheng Hou, Yuyu Luo, Hongyang Du (* equal contribution, ‡ project lead)

Preprint 2026

We present a systematic survey and unified taxonomy of efficient tool-using LLM agents, organizing techniques across tool context, calling, reasoning, and execution. We synthesize shared mechanisms, limitations, open challenges, and future directions for reducing resource use while preserving task performance.

[Paper]

How to Make Tool-Using LLM Agents Efficient? A Survey

Yunuo Hu*, Ziyang Cheng*, Wenqi Pei*‡, Miaomiao Li, Boyan Li, Yizhang Zhu, Yupeng Xie, Han Chen, Shizheng Hou, Yuyu Luo, Hongyang Du (* equal contribution, ‡ project lead)

Preprint 2026

We present a systematic survey and unified taxonomy of efficient tool-using LLM agents, organizing techniques across tool context, calling, reasoning, and execution. We synthesize shared mechanisms, limitations, open challenges, and future directions for reducing resource use while preserving task performance.

[Paper]

ROSE: An Intent-Centered Evaluation Metric for NL2SQL
ROSE: An Intent-Centered Evaluation Metric for NL2SQL

Wenqi Pei*, Shizheng Hou*, Boyan Li*, Han Chen, Zhichao Shi, Yuyu Luo (* equal contribution)

ACL 2026

We introduce ROSE, an intent-centered evaluation metric for NL2SQL that uses an adversarial Prover-Refuter cascade to judge whether predicted SQL captures a user's intent. ROSE aligns more closely with expert assessments and supports a large-scale re-evaluation of 19 NL2SQL methods.

[Paper] [Code]

ROSE: An Intent-Centered Evaluation Metric for NL2SQL

Wenqi Pei*, Shizheng Hou*, Boyan Li*, Han Chen, Zhichao Shi, Yuyu Luo (* equal contribution)

ACL 2026

We introduce ROSE, an intent-centered evaluation metric for NL2SQL that uses an adversarial Prover-Refuter cascade to judge whether predicted SQL captures a user's intent. ROSE aligns more closely with expert assessments and supports a large-scale re-evaluation of 19 NL2SQL methods.

[Paper] [Code]

NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions [Experiment, Analysis & Benchmark]
NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions [Experiment, Analysis & Benchmark]

Shizheng Hou, Wenqi Pei, Nuo Chen, Quang-Trung Ta, Peng Lu, Beng Chin Ooi

VLDB 2026

We present NL2SQLBench, the first modular benchmarking framework for LLM-enabled NL2SQL approaches. We dissect NL2SQL systems into three modules: Schema Selection, Candidate Generation, and Query Revision. We review existing strategies and propose fine-grained metrics that systematically quantify module-level effectiveness and efficiency.

[Paper] [Code]

NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions [Experiment, Analysis & Benchmark]

Shizheng Hou, Wenqi Pei, Nuo Chen, Quang-Trung Ta, Peng Lu, Beng Chin Ooi

VLDB 2026

We present NL2SQLBench, the first modular benchmarking framework for LLM-enabled NL2SQL approaches. We dissect NL2SQL systems into three modules: Schema Selection, Candidate Generation, and Query Revision. We review existing strategies and propose fine-grained metrics that systematically quantify module-level effectiveness and efficiency.

[Paper] [Code]

InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models with Human Feedback
InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models with Human Feedback

Hengyuan Zhao*, Wenqi Pei*, Yifei Tao*, Haiyang Mei, Mike Zheng Shou (* equal contribution)

EMNLP Finding & ICLR Bi-Align (Oral) 2025

This work introduces InterFeedback, a novel approach to understanding and enhancing the interactive intelligence of large multimodal models through human feedback mechanisms. Our framework provides new insights into model behavior and interaction patterns.

[Paper]

InterFeedback: Unveiling Interactive Intelligence of Large Multimodal Models with Human Feedback

Hengyuan Zhao*, Wenqi Pei*, Yifei Tao*, Haiyang Mei, Mike Zheng Shou (* equal contribution)

EMNLP Finding & ICLR Bi-Align (Oral) 2025

This work introduces InterFeedback, a novel approach to understanding and enhancing the interactive intelligence of large multimodal models through human feedback mechanisms. Our framework provides new insights into model behavior and interaction patterns.

[Paper]

Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models
Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models

Wenqi Pei, Hailing Xu, Hengyuan Zhao, Shizheng Hou, Han Chen, Zining Zhang, Pingyi Luo, Bingsheng He

IJCNLP-AACL Finding & ICLR DL4C 2025

We present Feather-SQL, a novel framework for Natural Language to Structured Query Language (NL2SQL) that leverages dual-model collaboration paradigm specifically designed for small language models. Our approach achieves state-of-the-art (SOTA) results on the BIRD benchmark.

[Paper] [Code]

Feather-SQL: A Lightweight NL2SQL Framework with Dual-Model Collaboration Paradigm for Small Language Models

Wenqi Pei, Hailing Xu, Hengyuan Zhao, Shizheng Hou, Han Chen, Zining Zhang, Pingyi Luo, Bingsheng He

IJCNLP-AACL Finding & ICLR DL4C 2025

We present Feather-SQL, a novel framework for Natural Language to Structured Query Language (NL2SQL) that leverages dual-model collaboration paradigm specifically designed for small language models. Our approach achieves state-of-the-art (SOTA) results on the BIRD benchmark.

[Paper] [Code]

All publications