HomeAboutResearchPublicationsProjectsBlogCVAppendixContact

All Posts

2026-08-24What Does It Mean to Remove an Object? From Image Inpainting to Video Object Removal2026-08-21Does Getting the Right Answer Improve Reasoning? The Promise and Limits of RLVR2026-08-20AI Does Not Remember the World — Approximation, Control, and Responsibility2026-08-17MCP and Harness Engineering: Building an Environment Where LLMs Can Work2026-08-11From SGD to Muon: Optimizer Families and Update Rules in 20262026-08-10Why Did the KV Cache Become an LLM Inference Bottleneck? From Architecture to TurboQuant2026-08-10Sovereign AI through the K-EXAONE Case — The Data Engine Comes Before the Model2026-08-07How Did SigLIP and SigLIP 2 Extend CLIP?2026-08-06Fundamental Concepts I Should Have Known2026-08-05Whose Model Can We Call Our AI? — Sovereign AI in South Korea2026-08-04Efficiently Batching Variable-Length Sequences for Transformers2026-08-04How Are LLMs Trained? — From Pre-training to Distillation and RL