The Feynman Technique: How to Use It to Actually Understand (Not Just Memorize)

You can recite the definition of backpropagation perfectly. Ask you to explain it to someone who's never heard of it, in plain language, and you stumble halfway through. That gap is exactly what the Feynman Technique is built to find.

It's named after physicist Richard Feynman, known for explaining famously difficult ideas in simple terms. The technique isn't really about teaching — it's a diagnostic tool that uses teaching to expose what you don't actually understand.

The four steps

1. Write the concept at the top of a page. Pick something specific — not "neural networks," but "why ReLU helps avoid the vanishing gradient problem." Specific concepts expose gaps faster than broad ones.

2. Explain it as if to someone with no background. Write it out, or say it out loud, using plain language — no jargon, no borrowed textbook phrases. Imagine a smart friend outside your field is listening. If you catch yourself using a technical term you haven't defined, that's already a sign.

3. Find where you got stuck. Somewhere in the explanation, you'll hit a point where you reach for a term you can't actually unpack, or the logic doesn't quite connect. That stumble is the entire point of the exercise — it's more useful than an hour of rereading, because it tells you precisely what to study next.

4. Go back to the source and simplify further. Return to your notes or textbook specifically for that gap. Once you understand it, rewrite your explanation, and try to simplify the language even more. If you still need jargon to explain a jargon term, you're not done yet.

Why it works better than it sounds

Explaining something out loud feels almost too simple to be a real study method. But it works because it forces active recall (producing the explanation from memory) combined with a strict test: plain language doesn't let you hide behind vocabulary you've memorized but not understood. Rereading lets vague understanding pass unnoticed. Explaining it simply doesn't.

Where it fits in a real study session

This isn't a replacement for reading your course material — it's what you do after a first pass, to find out what didn't actually land.

• After finishing a section, close your notes and try to explain the core idea in a couple of sentences, plain language only
• Do this before moving to flashcards or practice questions — it tells you where those tools should focus
• Keep a running list of concepts where you got stuck; that list is a better revision plan than "review everything"

A shortcut if you don't have anyone to explain it to

You don't need a real audience. Writing the explanation works almost as well as saying it out loud, and it leaves you a record to check later. Some students explain to a pet, a mirror, or just an empty room — the mechanism is the same either way: production under a plain-language constraint, not the presence of a listener.

Common mistake: explaining what you already understand well

It's tempting to Feynman-test the concept you already feel confident about, because it goes smoothly and feels validating. That's the least useful place to spend the time. Point this technique specifically at the topics you're least sure of — the ones you'd usually avoid testing, because that's exactly where it earns its keep.

Recap

The Feynman Technique works by forcing you to explain a concept in plain language, which exposes exactly where your understanding breaks down. Use it after a first read-through, on your weakest topics specifically, and let the sticking points — not a general reread — direct what you study next.

Kora lets you ask questions from your own course material when you hit a gap the Feynman technique reveals, so you're not stuck guessing where to look.