AtomWorld-Mem restores the hidden evolutionary state missing from instantaneous crystal snapshots, enabling physically grounded and transferable long-horizon atomistic evolution.
|
I am an undergraduate student at Sichuan University , pursuing dual bachelor's degrees in Mathematics and Artificial Intelligence. My research interests lie in AI for Science (AI4S), Large Language Models (LLMs), and scalable physical intelligence. Currently, I am keeping abreast of recent advances in LLM reasoning, generative AI, and other cutting-edge fields. I look forward to exploring potential research collaborations and pursuing PhD opportunities. Please feel free to contact me. Email: luotian0125@gmail.com | Google Scholar | CV | LinkedIn |
|
-
Our paper AtomWorld-Mem was accepted to NeurIPS 2026! It is my first research paper and was accepted on my first submission. Special thanks to Kun, who introduced me to research and has had a profound influence on my journey.
-
I received the National Scholarship for the second consecutive year! I am grateful for my hard work and for the support of my family, friends, teachers, and everyone who has helped me grow.
More newsLess news1 older update
-
I received the National Scholarship and Comprehensive First-Class Scholarship of Sichuan University as the sole recipient in my major.
† Equal contribution. ‡ Corresponding author.
Understanding material degradation over engineering timescales remains a fundamental challenge. We introduce AtomWorld, a physics-grounded atomistic world model that restores hidden kinetic context to bridge atomic events and material degradation across metre and year scales.
|
Sichuan University
2024–2028 (Expected) Undergraduate Student Dual bachelor's degrees in Mathematics and Artificial Intelligence |
|
Institute for AI Industry Research (AIR), Tsinghua University
Feb. 2026 – Oct. 2026 Research Intern Research Advisor: Assistant Professor Kun Li |
- 2025, 2026 — National Scholarship
- 2025 — Comprehensive First-Class Scholarship, Sichuan University
