I’m an AI safety researcher — mostly working on ways to see inside the systems we’ve built and understand what moves them. My background runs through statistical inference, machine learning, and generative models; lately I’ve been in the borderlands between mechanistic interpretability and probabilistic thinking, trying to make large language models a little less opaque.
I’ve moved between academia, industry, and independent research, but the constant thread is the same: bridging abstract mathematics with the hidden structures of deep networks. I view this blog as the place to keep scientific diary, maintain emotional balance, and make friends :)
Indeed lots of wordings and sec arrangement in the first version, just got over-excited and wanted to record the new great algorithm fast, not expecting early readers, not sure about language.
Anyway just changed bit now and hope it’s easier to chew. Everything else is still 100% original and poured loads of effort, just feel very shocked of the votes and these parts ignored.