🎓 PhD Student @ Aalto University
💡 Researching Large Output Space and Efficient LLM
🧑💻 Former R&D Engineer (Search Engine and Pre-trained Language Models) at two tech companies
✨ Passionate about efficient deep learning systems
🧑🏫 Working with Prof. Rohit Babbar and Prof. Pekka Marttinen
⚙️ Currently learning and developing with CUDA
🎯 Goal: Completing my PhD and starting a side project on ML systems in the near future
🧠 Current Focus: Accelerating LLM inference by optimizing large output spaces with CUDA stream parallelism, and exploring 8-bit / 4-bit low-precision training for efficient model scaling.
📄 Recent Publication: "ELMO: Efficiency via Low-precision and Peak Memory Optimization in Large Output Spaces" — ICML 2025.
- Large Output Space and Extreme Classification
- Efficient LLM Inference and Training
- Low-Precision Training (8-bit / 4-bit)
- GPU Programming (CUDA, Trion)
💼 LinkedIn