Undergraduate Researcher · Shanghai Jiao Tong University
Kaicheng Yang
I work between efficient machine learning and AI4Human: compressing models to identify the necessary structures intelligence truly depends on, and using AI to understand the human brain while expanding the interaction bandwidth between carbon-based and silicon-based intelligence.
Efficient MLI use model compression as a scientific probe: by stripping away redundancy through quantization, sparsity, and low-rank structure, I try to identify the necessary structures that intelligence truly depends on and use them to move toward next-generation AI architectures.
AI4HumanI want to use AI to understand the human brain, explore the next generation of human-computer interaction, and increase the communication bandwidth between carbon-based and silicon-based intelligence.
News
2026.05 · RobuQ, Info-Gain, and Q-DiT4SR accepted to ICML 2026.
2026.02 · Three preprints released on arXiv.
2026.01 · PT^2-LLM accepted to ICLR 2026.
2025.10 · National Challenge Cup Grand Prize.
2025.09 · NSFC undergraduate research project support.
2025.09 · Three preprints released on arXiv.
2025.05 · BiMaCoSR accepted to ICML 2025.
Honors
- 2025 · National Challenge Cup Grand Prize
- 2025 · NSFC Undergraduate Research Project Support
- 2024 · Infineon Scholarship
- 2024 · University Second-Class Scholarship
- 2023 · First Prize, National College Student Mathematics Competition