What is Reranking?Reranking kya hai?

🧸 Think of it likeAise samjho

Like shortlisting 50 job applicants quickly, then interviewing them carefully to pick the best 5.Jaise 50 applicants ko jaldi shortlist karna, phir dhyaan se interview karke best 5 chunna.

📘 ExplainedVistaar se

First retrieval quickly finds, say, 50 candidates; a reranker model reads the question and each candidate together and orders them by true relevance, keeping the best few for the LLM. It adds some latency but usually improves RAG answers a lot.Pehla retrieval jaldi, maan lo, 50 candidates dhoondhta hai; reranker model sawaal aur har candidate ko saath padh kar asli relevance se order karta hai, aur LLM ke liye kuch best rakhta hai. Thodi latency judti hai par aksar RAG jawaab kaafi behtar hote hain.

💡 Real exampleAsli example

Reordering 50 search hits so the exact policy clause moves from position 12 to position 1.50 search hits ko aise reorder karna ki exact policy clause position 12 se 1 pe aa jaaye.

📖 Learn it step by stepStep by step seekho

• RAG: letting an LLM answer from your documentsRAG: LLM ko aapke documents se jawaab dene dena

🔗 Related termsJude hue shabd

• Hybrid searchCombining keyword search and vector search to get the best of both.Keyword search aur vector search ko milana taaki dono ka faayda mile.
• RAGLetting an LLM answer from your documents: search for relevant passages first, then generate an answer from them with citations.LLM ko aapke documents se jawaab dene dena: pehle relevant hisse dhoondho, phir unhi se citations ke saath jawaab banao.
• Lost in the middleLLMs tend to use information at the start and end of a long prompt better than information in the middle.LLMs lambe prompt ki shuru aur aakhir ki information ko beech wali se behtar use karte hain.