hAIgopala › AI terms › Rapid Fire (3 terms per short) › Day 16

Scaling Laws

Generative AI & LLMsChapter 4: Choosing the LLMTerm 78 of 253

Performance predictably improves with more data, parameters and compute.

At Sunrise School

Gopal asks

Why is each new model smarter?

Bit answers

Scaling laws: more data, parameters and compute predictably improve quality. That's why newer models handle tricky enquiries better.

In the real world

Bigger engine + more fuel = faster car (to a point).

Watch the short

Short coming soonSeason 5 · Day 16Follow @haigopala

Taught together with

Instruction Tuning · Constitutional AI / RLAIF

Practise it

Try this term and all 253 others in the free Model Training Arena.

Open the Arena →

Follow Bit daily: YouTube English · తెలుగు · हिंदी  |  Instagram English · తెలుగు · हिंदी