Fundamental Limitation in Explaining AI

Atsushi Suzuki, Jing Wang

While large-scale models such as LLMs and diffusion models have achieved practical success, public institutions have emphasized the importance of explainability in AI. Existing met…