AI dictionary · LLMs and GenAI
What is RLHF?RLHF kya hai?
Also called: Reinforcement learning from human feedback
RLHF: Training an LLM to prefer answers that people rate as better.LLM ko aise jawaab pasand karna sikhaana jinhe log behtar rate karte hain.
🧸 Think of it likeAise samjho
Like a trainee chef cooking two versions, a panel of tasters picking the better one, and the chef slowly learning what people like.Jaise trainee chef do version banaye, tasters ka panel behtar wala chune, aur chef dheere dheere seekhe ki logon ko kya pasand hai.
📘 ExplainedVistaar se
People compare pairs of model answers; a reward model learns these preferences; the LLM is then optimised with reinforcement learning to produce higher rewarded answers. It improves helpfulness and safety. DPO is a simpler alternative that skips the separate reward model.Log model ke jawaabon ke jodon ko compare karte hain; reward model yeh pasand seekhta hai; phir LLM ko reinforcement learning se zyaada reward wale jawaab dene ke liye optimise kiya jaata hai. Isse helpfulness aur safety sudharti hai. DPO simple alternative hai jo alag reward model chhod deta hai.
💡 Real exampleAsli example
Raters prefer a clear, polite answer over a rambling one; the model learns that style.Raters lambi baaton ke bajaye saaf, vinamra jawaab pasand karte hain; model woh style seekhta hai.
📖 Learn it step by stepStep by step seekho
🔗 Related termsJude hue shabd
Questions people ask
What is RLHF?
Training an LLM to prefer answers that people rate as better. People compare pairs of model answers; a reward model learns these preferences; the LLM is then optimised with reinforcement learning to produce higher rewarded answers. It improves helpfulness and safety. DPO is a simpler alternative that skips the separate reward model.
How can I understand RLHF in simple words?
Like a trainee chef cooking two versions, a panel of tasters picking the better one, and the chef slowly learning what people like.
What is an example of RLHF?
Raters prefer a clear, polite answer over a rambling one; the model learns that style.
RLHF ka matlab kya hai?
LLM ko aise jawaab pasand karna sikhaana jinhe log behtar rate karte hain. Jaise trainee chef do version banaye, tasters ka panel behtar wala chune, aur chef dheere dheere seekhe ki logon ko kya pasand hai.
See RLHF in the full A to Z dictionary · Start a free AI course