I'm at #NeurIPS 2025 and will present our paper REASONING COMPILER: LLM-Guided Optimizations for Efficient Model Serving!
Our work uses LLM reasoning to compile LLMs more efficiently and with higher sample efficiency, creating a virtuous cycle of LLMs improving themselves.
Poster Location: Exhibit Hall C/D/E, Booth: #909
Time: Fri, Dec 5 4:30 PM — 7:30 PM PST
NeurIPS: https://t.co/8fFyFLiZQH
Hope to see you there!
Shout out to our great team members: Hadi Esmaeilzadeh(@catsanroses), Lianhui Qin(@Lianhuiq), Christopher Priebe, and Rohan Mahapatra
#NeurIPS2025 #LLM #LLMReasoning #LLMServing #GenAI #AI #AIInnovation #ML #DeepLearning #Compiler #OpenSource