Nanjing University
Shihao Li
Master’s student in artificial intelligence
I’m pursuing a master’s degree in artificial intelligence at Nanjing University’s School of Intelligence Science and Technology, where I work with my advisor, Professor Jiaheng Liu.
My earlier research focused on evaluating multimodal and omni-modal large models. I now focus on personal AI agents and AI education.
I earned my bachelor’s degree at Shanghai University’s School of Computer Engineering and Science (2021–2025).
NJU-LINK
Large-scale Intelligence and Knowledge Group
Publications
First or co-first author
-
IF-VidCap: Can Video Caption Models Follow Instructions?
ICLR 2026
-
ViDiC: Video Difference Captioning
NeurIPS 2026
-
MMG2Skill: Can Agents Distill In-the-Wild Guides into Self-Evolving Skills?
EMNLP 2026 Findings
Core contributor
-
OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs
ICLR 2026
Collaborative papers
-
OmniHalluc-L: Counterfactual Benchmarking and Modality-Perturbation Calibration for Long-Form Omni Hallucination
EMNLP 2026
-
DR³-Eval: Towards Realistic and Reproducible Deep Research Evaluation
EMNLP 2026
-
CapRiCorn-1K: A Comprehensive Benchmark for Video Captioning and Subject Referential Consistency Across Temporal Scales
EMNLP 2026
-
MT-Video-Bench: A Holistic Video Understanding Benchmark for Evaluating Multimodal LLMs in Multi-Turn Dialogues
ACL 2026 Findings
-
OmniBench: Towards The Future of Universal Omni-Language Models
NeurIPS 2025
-
MVU-Eval: Towards Multi-Video Understanding Evaluation for Multimodal LLMs
NeurIPS 2025
Preprints
-
Vibe AIGC: A New Paradigm for Content Generation via Agentic Orchestration
Preprint · arXiv 2026