[visarga]

> The most important criteria I look for when I interview applicants is what they have built.

How do you know the extent of their contribution? Maybe the project was cool but the applicant only had a marginal participation.

I prefer to ask a few comprehension questions that focus on concepts we use every day on the job, like what's the cost of doubling the sequence length for GPT models, or what is cosine similarity and its applications. They should demonstrate they can operate in this space and have a good grasp on the basics.

[amayne]

I look at their personal GitHub repos (when they're relevant.)