Open call

The Institute funds undergraduates doing theoretical work at the intersection of learning theory, statistical physics and computation. The money is modest and the point is not the money: it is that a student with a real question should be able to spend a summer on it instead of on something that pays better.

What it is for

A specific question you want to answer, and the concrete costs of answering it — compute time, a stipend that lets you work on this rather than take other work, travel to the one workshop where the people who care about your problem will be, the page charges on a preprint.

We would rather fund one clearly stated question than a broad programme of study. A proposal that names the thing it is trying to find out, and says what would count as having found it out, is already most of the way there.

Who can apply

• StandingCurrently enrolled as an undergraduate, anywhere. No requirement to be at a research university — if anything, the grant is worth more to someone who is not.
• FieldMathematics, physics, computer science, or something that does not have a department yet. The subject has to touch the Institute's interests; your degree title does not have to.
• SupervisionAn advisor is welcome but not required. We would rather read a good proposal from a student with no one to sign it than a routine one with a signature.
• Prior workNo publications expected. A half-finished attempt that taught you where the difficulty actually is counts for more than a clean record.

What to send

One message, in plain prose, covering four things: the question you want to work on and why it is not already answered; how you would attack it, concretely enough that we can tell whether it would work; what you would need and roughly what it costs; and who you are — where you study, what you have taught yourself, what you have tried.

No form to fill in beyond the fields below, no reference letters, no transcript. If we want more we will ask. Links to anything you have written or built are the most useful thing you can include.

How it is decided

Read by the people who would be working with you, against one question: is there a real question here, and would this money let this person get at it? Everyone who applies gets an answer. Applications are read as they arrive rather than in rounds, so there is no advantage to waiting and none to rushing.

Apply

This goes to the same place as everything else sent to the Institute, tagged as a grant application so it does not get lost among correspondence.