How to write a literature review that actually gets cited
PhD @ Stanford · ML researcher
Most literature reviews read like annotated bibliographies. Here's how to write one that adds real value to your field.
A great literature review tells a story. It identifies themes, contradictions, and gaps. It doesn't just summarize — it synthesizes.
My framework:
1. Map the landscape — organize papers by methodology, not chronology
2. Identify the debates — where do researchers disagree?
3. Find the gaps — what hasn't been studied yet?
4. Connect to your work — why does this gap matter?
The best literature review I ever read was 8 pages long and cited 120 papers. Every sentence earned its place.
Comments (12)
This is exactly what I needed. The part about communication being more important than the perfect design really resonated with me.
Just passed my system design round using these tips. The requirements clarification step saved me from going down the wrong path.
Thanks everyone! Happy to answer any follow-up questions about the process.
Absolutely go for it! The worst case is you get great practice. I failed my first attempt and learned a ton from it.
The tip about explaining designs to non-technical friends is gold. I started doing this and my communication improved dramatically.
I'd add one more thing: always discuss monitoring and observability. Interviewers love when you think about production readiness.
How long did you spend on each mock interview session? I find it hard to simulate the real pressure.
Sharing this with all my mentees. Concise and actionable — exactly what interview prep content should be.