Learner, Developer, and Researcher. I build multimodal foundation models and enjoy turning research ideas into things that actually run.
🔭 I'm an AI Scientist at LG AI Research (EXAONE Lab), working on EXAONE 5.0 (a MoE multimodal LLM with image & video understanding) and the next-generation sovereign model K-EXAONE 3, focusing on Video & Document Understanding. Previously contributed to EXAONE 4.5. My team also developed K-EXAONE and K-EXAONE-2 — and we're always hiring.
🎓 I earned my Ph.D. in Computer Science from Yonsei University, advised by Prof. Seon Joo Kim, on efficient generative vision. Previously interned at Adobe Research and NAVER.
- Multimodal large language models (image & video understanding)
- Generative vision models (Diffusion, GAN, NeRF) and face modeling
- Vision–language & vision–audio representation learning
Python · PyTorch · Megatron · Diffusers · Docker · GCP
- 🌐 Homepage: yangspace.co.kr
- 📝 Blog: blog.yangspace.co.kr
- 🎓 Google Scholar: profile
- 💼 LinkedIn: sejong-yang
For CV, publications, and projects, see my homepage.