All Posts
What Does It Mean to Remove an Object? From Image Inpainting to Video Object RemovalDoes Getting the Right Answer Improve Reasoning? The Promise and Limits of RLVRAI Does Not Remember the World — Approximation, Control, and ResponsibilityMCP and Harness Engineering: Building an Environment Where LLMs Can WorkFrom SGD to Muon: Optimizer Families and Update Rules in 2026Why Did the KV Cache Become an LLM Inference Bottleneck? From Architecture to TurboQuantSovereign AI through the K-EXAONE Case — The Data Engine Comes Before the ModelHow Did SigLIP and SigLIP 2 Extend CLIP?Fundamental Concepts I Should Have KnownWhose Model Can We Call Our AI? — Sovereign AI in South KoreaEfficiently Batching Variable-Length Sequences for TransformersHow Are LLMs Trained? — From Pre-training to Distillation and RL