Get the latest tech news
LLMs understand nullability
_{model} The last five years have shown us that large language models, like ChatGPT, Claude, and DeepSeek, can write code in many domains, to huge excitement: many claim to be using these models to write entire web servers and apps from scratch. These tools have opened up programming to a whole new class of people who consider themselves non-technical.
This makes sense, since some of the most critical LLM tasks involve chatting with a user, and some of the most interesting concepts to measure, such as honesty or power-seeking, apply most readily to these conversations. If we pick the right properties, we don’t need to worry about our ability to label data— static analysis can do that for us, and so we can easily scale up and train on thousands of examples generated from scratch. At this point, we’ve figured out how to roughly measure nullability understanding in the output of various language models, but we still don’t know what their internal representations might look like or when they emerge.
Or read this on Hacker News