LLM: “Here’s an optimized SQL query.”
Me: "EXPLAIN ANALYZE"
The query: 💀
One habit that makes LLM-generated SQL 10x better:
"EXPLAIN ANALYZE"
Don't just ask the LLM if the query is optimized. Let the database tell you.
One of the best C pointer exercises from MIT.
I challenge you to predict the full output without running the code.
This is real pointer stuff. Not the basic ones.
Save it. Try it when you think you know C well.
Code here:
https://t.co/u9KqqGqMtj
RAG fundamentals:
1/ Why do "advanced" retrieval techniques like HyDE and MultiQuery sometimes underperform plain dense retrieval? Full benchmark and lessons in this thread.
RAG fundamentals: The HyDE hallucination paradox
1/ Why did adding "advanced" retrieval like HyDE drop your Hit Rate instead of lifting it? You expected a boost. Hit Rate fell 8 points.
1/ Have you wondered why your RAG works in demos but fails on real queries?
Chunk boundary failure could be the cause.
Three fixes for the most overlooked retrieval failure: chunk boundary splits.
Ever wondered what magic happens behind the scenes when you type a command into your shell and hit enter? 🤔 Just wrapped up a deep dive into the inner workings of the shell! 🚀 Dive in with me and unravel the mystery: https://t.co/HKf7G5E1ju