I'm Nathan Axcan, current intern @ IBM Research Zurich
- (systems work) implementing "span semantics" a.k.a. parallel prefill, independent caching, etc on IBM's official vllm fork
- (DL research work) researching independent caching opportunities of hybrid-architecture LLMs
- (DL foundational work) researching rendered LLMs
Three things about me:
- I love to talk about interesting code, ideas, dig deep, think as deep as humanly possible
- Love the entrepreneurial/economics/incentives aspects of most topics
- In love with technology, philosophy, but my biggest love of all is deep learning
Shoot an email over at nathanaxcan@gmail.com and I'll be happy to hear what you have to say. Who knows, maybe we can help eachother out!
- startup analysis @Plug&Play Tech Center
- messing around in python (finetuned a chatbot before ChatGPT came out)
- former data scientist @Cortexter
- podcasted about motivation and productivity @Slim Radio
- co-led development for EU-funded VR + holograms ed-tech research project Hololearn
- strategic advising (startups, projects)