Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Yixuan Li

Yixuan Li

3 indexed papers

Recent (6 mo)
3
With code
0
Influential cites
0
Benchmarked
0

Publications per year

3
26

Top categories

Vision×2Software Eng.×1AI×1Sound×1

Frequent co-authors

Yin Wu1×
Yixuan Liu1×
Yi Li1×
Chenyang Peng1×
Hao Wu1×
Ming Fan1×

Research Timeline

2026
video-SALMONN-R$^3$: Learning to ReWatch, ReAsk, and ReAnswer for Efficient Video Understanding

This paper presents video-SALMONN-R$^3$, an end-to-end video-LLM that enables re-watch through reinforcement learning, improving question answering performance with a two-stage paradigm.

WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

The paper introduces WorldDirector, a framework for creating controllable video worlds with persistent dynamic object memory and exact visual identities.

TrapHunter: Exposing Covert Pathways in Trap Token Contracts

This paper proposes TrapHunter, a framework to identify deceptive trap tokens in standardized contracts using intent deviation analysis and dynamic validation.

Highlighted terms show continued research focus across papers

Papers

cs.SEEmpiricalRecentJul 21, 2026

TrapHunter: Exposing Covert Pathways in Trap Token Contracts

Yin Wu, Yixuan Liu, Yi Li, Chenyang Peng +4 more

This paper proposes TrapHunter, a framework to identify deceptive trap tokens in standardized contracts using intent deviation analysis and dynamic validation.

View →
cs.CVEmpiricalRecent
Jul 2, 2026

WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

Hanlin Wang, Hao Ouyang, Qiuyu Wang, Wen Wang +9 more

The paper introduces WorldDirector, a framework for creating controllable video worlds with persistent dynamic object memory and exact visual identities.

View →
cs.CVcs.AIcs.SDEmpiricalRecentJun 23, 2026

video-SALMONN-R$^3$: Learning to ReWatch, ReAsk, and ReAnswer for Efficient Video Understanding

Yixuan Li, Guangzhi Sun, Yudong Yang, Wei Li +2 more

This paper presents video-SALMONN-R$^3$, an end-to-end video-LLM that enables re-watch through reinforcement learning, improving question answering performance with a two-stage paradigm.

View →