Team Member

Zhengxin Yang

Founder

Employee Current Employee

In Lab: 2024-09-01 - Present

Works on AIDAI and Evaluation.

Email: yangzhengxin@ict.ac.cn Website Updated: 2026-03-17

Profile

Zhengxin Yang is the founder of YangsAI, working on the design, evaluation, and systematization of artificial intelligence systems.

His research focuses on Artificial Intelligence Designed Artificial Intelligence (AIDAI), which studies how AI systems can design, evaluate, and improve other AI systems in a structured and iterative manner.

For a full academic CV, publications, and background, please refer to: 👉 https://jason-young.me

Research Interests

Artificial Intelligence Designed Artificial Intelligence (AIDAI)
AI Evaluation and Benchmarking

Selected Publications

Browse Publications

Defaulting to highlighted publications authored by Zhengxin Yang; you can switch to all publications below (top 6).

Browsing Publications

Scope All Year All Type All Venue All

Year Type Venue Publications 6 Publications

2026

Paper

When AI Designs AI: Innovation or Imitation?

Paper

Preprint | Aug 18, 2026

Yikang Yang First Zhengxin Yang Corr. Me Luzhou Peng Minghao Luo Yanqi Kan Wanling Gao Jianfeng Zhan

First:

Yikang Yang

Corresponding:

Zhengxin Yang

Preprint

When AI Designs AI: Innovation or Imitation?

@misc{yang2026aidesignsaiinnovation,
      title={When AI Designs AI: Innovation or Imitation?}, 
      author={Yikang Yang and Zhengxin Yang and Luzhou Peng and Minghao Luo and Yanqi Kan and Wanling Gao and Jianfeng Zhan},
      year={2026},
      eprint={2608.17471},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2608.17471}, 
}

Beyond Benchmarks: Toward Causally Faithful Evaluation of Large Language Models

Paper

International Conference on Machine Learning (ICML) | May 01, 2026

Zhengshuyuan Tian First Wanling Gao First Corr. Chuanxin Lan Chenxi Wang Lei Wang Guoxin Kang Zhengxin Yang Me Yunyou Huang Xuehai Hong Corr. Jianfeng Zhan

First:

Zhengshuyuan Tian , Wanling Gao

Corresponding:

Wanling Gao , Xuehai Hong

Paper

Beyond Benchmarks: Toward Causally Faithful Evaluation of Large Language Models

@inproceedings{
tian2026beyond,
title={Beyond Benchmarks: Toward Causally Faithful Evaluation of Large Language Models},
author={Zhengshuyuan Tian and Wanling Gao and Chuanxin Lan and Chenxi Wang and Lei Wang and Guoxin Kang and Zhengxin Yang and Yunyou Huang and Xuehai Hong and Jianfeng Zhan},
booktitle={Forty-third International Conference on Machine Learning},
year={2026},
url={https://openreview.net/forum?id=XAp0VUy20q}
}

GraDE: A Graph Diffusion Estimator for Frequent Subgraph Discovery in Neural Architectures

Paper

Preprint | Feb 03, 2026

Yikang Yang First Zhengxin Yang Corr. Me Minghao Luo Luzhou Peng Hongxiao Li Wanling Gao Lei Wang Jianfeng Zhan

First:

Yikang Yang

Corresponding:

Zhengxin Yang

Preprint

GraDE: A Graph Diffusion Estimator for Frequent Subgraph Discovery in Neural Architectures

@misc{yang2026gradegraphdiffusionestimator,
      title={GraDE: A Graph Diffusion Estimator for Frequent Subgraph Discovery in Neural Architectures}, 
      author={Yikang Yang and Zhengxin Yang and Minghao Luo and Luzhou Peng and Hongxiao Li and Wanling Gao and Lei Wang and Jianfeng Zhan},
      year={2026},
      eprint={2602.03257},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2602.03257}, 
}

Probing Memes in LLMs: A Paradigm for the Entangled Evaluation World

Paper

Preprint | Feb 03, 2026

Luzhou Peng First Zhengxin Yang Corr. Me Honglu Ji Yikang Yang Fanda Fan Wanling Gao Jiayuan Ge Yilin Han Jianfeng Zhan

First:

Luzhou Peng

Corresponding:

Zhengxin Yang

Preprint

Probing Memes in LLMs: A Paradigm for the Entangled Evaluation World

@misc{peng2026probingmemesllmsparadigm,
      title={Probing Memes in LLMs: A Paradigm for the Entangled Evaluation World}, 
      author={Luzhou Peng and Zhengxin Yang and Honglu Ji and Yikang Yang and Fanda Fan and Wanling Gao and Jiayuan Ge and Yilin Han and Jianfeng Zhan},
      year={2026},
      eprint={2603.04408},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2603.04408}, 
}

TimeMosaic: Temporal Heterogeneity Guided Time Series Forecasting via Adaptive Granularity Patch and Segment-wise Decoding

Paper

AAAI Conference on Artificial Intelligence (AAAI) | Feb 01, 2026

Kuiye Ding First Fanda Fan Corr. Chunyi Hou Zheya Wang Lei Wang Zhengxin Yang Me Jianfeng Zhan

First:

Kuiye Ding

Corresponding:

Fanda Fan

Paper

TimeMosaic: Temporal Heterogeneity Guided Time Series Forecasting via Adaptive Granularity Patch and Segment-wise Decoding

@inproceedings{ding2026timemosaic,
      title={TimeMosaic: Temporal Heterogeneity Guided Time Series Forecasting via Adaptive Granularity Patch and Segment-wise Decoding},
      author={Kuiye Ding and Fanda Fan and Chunyi Hou and Zheya Wang and Lei Wang and Zhengxin Yang and Jianfeng Zhan},
      booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
      year={2026}
    }

2024

Paper

Evaluatology: The Science and Engineering of Evaluation

Paper

BenchCouncil Transactions on Benchmarks, Standards and Evaluations (TBench) | Apr 30, 2024

Jianfeng Zhan First Lei Wang Wanling Gao Hongxiao Li Chenxi Wang Yunyou Huang Yatao Li Zhengxin Yang Me Guoxin Kang Chunjie Luo Hainan Ye Shaopeng Dai Zhifei Zhang

First:

Jianfeng Zhan

Paper

Evaluatology: The Science and Engineering of Evaluation

@article{ZHAN2024100162,
title = {Evaluatology: The science and engineering of evaluation},
journal = {BenchCouncil Transactions on Benchmarks, Standards and Evaluations},
volume = {4},
number = {1},
pages = {100162},
year = {2024},
issn = {2772-4859},
doi = {https://doi.org/10.1016/j.tbench.2024.100162},
url = {https://www.sciencedirect.com/science/article/pii/S2772485924000140},
author = {Jianfeng Zhan and Lei Wang and Wanling Gao and Hongxiao Li and Chenxi Wang and Yunyou Huang and Yatao Li and Zhengxin Yang and Guoxin Kang and Chunjie Luo and Hainan Ye and Shaopeng Dai and Zhifei Zhang}
}