Published papers include CoTAM@ACL2024F, Incubator@EMNLP2024, TextGraft@EMNLP2024, NDD@ACL2023F, X-NER@EMNLP2023F, MetaIE, Cuckoo@ACL2025, InBedder@ACL2024, etc.
Research Experience
2023-2024: Worked on building NLP systems with extremely weak signals from humans with collaboration from LLMs (Knowledge Discovery).
2024-2025: Worked on unveiling the mechanisms behind LLM’s memorization, diversity, generalization, and hallucination (Knowledge Parameterization).
2025-: Now working on fine-grained control of LLM’s roles in AI4Plot, AI4Game, etc.
Education
Currently a 4th-year Ph.D. student at UC San Diego, advised by Professor Jingbo Shang; B.S. degree in Computer Science from SJTU, advised by Professor Hai Zhao.
Background
Research interests include knowledge discovery, parameterization, and generation of large language models. Aiming to build knowledgeable LLMs (especially as experts in unique domains) with diversity, generalizability, and fine-granularity.
Miscellany
The personal homepage quotes Japanese lyrics “ここから始めよう♪ ボクは旅に出よう♪”, suggesting the author's enthusiasm for new journeys or challenges.