Fangru Lin

DPhil student at University of Oxford


Oxford, United Kingdom

Hi! I am Fangru and welcome to my page!

I’m reading for a DPhil degree at University of Oxford with a research focus on NLP. I’m working on linguistically informed LLMs, mainly on semantic/pragmatic implications on NLP technologies. I also want to know how human cognitive models can help us understand/develop LLMs.

I am supervised by Prof Janet Pierrehumbert and Prof Daniel Altshuler. My academic interests mainly lie in:

· Developing better models to understand complex linguistic meanings and reasoning paths for decision making and language understanding
· Very interested in neuro-symbolic methods and formal theories!

For most recent publications, please visit my Google scholar page.

I co-organize OxNLP, a hybrid weekly reading group sharing sota NLP works at University of Oxford. See more information and feel free to join our mailing list here.

My non-academic interests are in music and sports! I have level-10 (highest level) piano player certificate. I have also been in many sing comps including Voice of China, and Chinese sing comp (唱响津城) at University of Oxford (will be in more sing comps hopefully!) For sports, I often play golf and do Pilates. I sometimes ski.

See my other profile pages in Oxford e-Research Centre and Faculty of Linguistics, Philology, and Phonetics.


May 2, 2024 Our paper Graph-enhanced Large Language Models in Asynchronous Plan Reasoning is accpeted at ICML-2024! See you in Vienna!
Apr 21, 2024 I am researching LLMs at Microsoft Research. Looking forward to building something really cool!
Feb 21, 2024 I will be a panelist for “Large Language Models and Artificial General Intelligence” in OxAI mini-conference on 25th Feb. Join us if you are in Oxford!
Feb 20, 2024 My master thesis Probing Large Language Models for Scalar Adjective Lexical Semantics and Scalar Diversity Pragmatics is accepted at LREC-COLING 2024! Look forward to seeing people in Torino!
Feb 19, 2024 I will be giving a talk in MilaNLP about Graph-enhanced Large Language Models in Asynchronous Plan Reasoning under the theme of coding aperitivo on 29th March. Stay tuned!