About
Hello ☺️ I'm a student and a xcpcer / oier;
Since childhood, I have been enthusiastic about algorithmic problem solving and have devoted significant time to competitive programming, including OI and XCPC.
I am still exploring my long-term research interests.
And now I'm happy that I'm learning the Agentic RL and Learning Theory with Prof. JianQian.
In prior time, I was fortunate to study Computer Vision under the guidance of Prof. Shaodian and learned some knowledge of RANLA 4 ML under Prof. Liao Zhenyu. Through these experiences, I started to reflect more carefully on the mathematical structures and theoretical foundations behind modern AI systems.
You can see my CV here and my Full Award here . Feel free to contact me at yinxinxiang24@gmail.com to discuss anything.
Selected Publications
View All →Show, Don't Tell: Morphing Latent Reasoning into Image Generation
Xinxiang Yin†, Haodong Chen†, Wen-Jie Shu, Hongfei Zhang, Zixin Zhang, Litao Guo, Qifeng Chen, Ying-Cong Chen
arXiv preprint
A novel framework that seamlessly integrates implicit latent reasoning into the T2I generation process, enhancing Janus-Pro by 16% on GenEval while reducing inference time by 44%.
CodeHacker: Automated Test Case Generation for Detecting Vulnerabilities in Competitive Programming Solutions
Jingwei Shi†, Xinxiang Yin†, Jing Huang†, Jinman Zhao, Shengyu Tao
ACL main conference
Accepted at ACL 2026 Main Conference. An automated agent framework dedicated to generating targeted adversarial test cases that expose latent vulnerabilities in program submissions, improving the True Negative Rate of existing datasets.
FineQuest: Adaptive Knowledge-Assisted Sports Video Understanding via Agent-of-Thoughts Reasoning
Haodong Chen, Haojian Huang, Xinxiang Yin, Dian Shao
ACM Multimedia oral presentation
Oral presentation at ACM MM 2025. The first training-free framework for sports video understanding that leverages dual-mode reasoning inspired by cognitive science.
News
Our work CodeHacker is accepted at ACL 2026 Main Conference!
Our work LatentMorph has been reported by public media.
Qualified for 50th ICPC World Finals in Hangzhou. Submitted two papers for ACL'26 and ICML'26.
Achieved Codeforces Grandmaster rank.
Won gold medals at ICPC Shenyang, ICPC Shanghai (10th place), and CCPC Chongqing.
Won 2nd place at International Advanced Robot & Simulation Technology Competition in Qingdao.
Our paper FineQuest was accepted at ACM MM 2025 as oral presentation.
Won Shaanxi Provincial Programming Contest Championship.