How do I get better answers?
Nine habits, explained: you know what the model actually sees, why it ignores or invents things, and how to keep it fast, cheap and on script.
- 1Why does it forget what I said 20 messages ago?
Context window · Beginner · 3 min
- 2Why does it ignore the middle of my long document?
Lost in the middle: position bias in long contexts · Beginner · 5 min
- 3Why does the model ignore my instructions?
System prompts and instruction priority · Beginner · 5 min
- 4Why is the second call cheaper?
Prompt caching · Builder · 5 min
- 5Why does the free tier say 429?
Rate limits (RPM, TPM, RPD) · Builder · 5 min
- 6Why does it break my JSON?
Structured output (constrained decoding) · Builder · 5 min
- 7Why can a web page hijack my AI agent?
Prompt injection · Builder · 5 min
- 8Why does it invent sources that don't exist?
Hallucination · Beginner · 3 min
- 9Why doesn't it know about last week?
Knowledge cutoff: training data stops at a date · Beginner · 5 min
- 10Why does it miscount the r’s in "strawberry"?
coming · opens from the tokens station
- 11How does it know "king" is close to "queen"?
coming · opens from the embeddings station
- 12How does it know what "it" refers to?
coming · opens from the attention station
- 13What's inside the box?
coming · opens from the layers station
- 14Does it plan what it's going to say?
coming · opens from the sampling station
- 15Why does the same prompt give different answers?
coming · opens from the sampling station
- 16How did it learn all this?
coming · opens from the layers station
- 17Why is it so polite (and so cautious)?
coming · opens from the layers station
- 18Teach it my docs, or let it read them?
coming · opens from the embeddings station
- 19What’s an "agent", and why is everyone saying it?
coming · opens from the play station
- 20Why does it ‘think’ before answering?
coming · opens from the sampling station
- 21Why does the free model feel dumber?
coming · opens from the layers station
- 22How does a 30 GB model fit on my laptop?
coming · opens from the layers station
- 23How does it read a picture?
coming · opens from the tokens station
- 24How does AI draw a picture from noise?
coming · opens from the sampling station
- 25Why does the answer appear word by word?
coming · opens from the sampling station
- 26Why is the mini or flash model the free one?
coming · opens from the layers station
- 27Why can I run Llama on my laptop but not GPT?
coming · opens from the layers station
- 28Why does a 7B model need 16 GB of RAM?
coming · opens from the layers station
- 29Why does Vietnamese cost more tokens than English?
coming · opens from the tokens station
- 30Why is DeepSeek so cheap (and why is a 671B model fast)?
coming · opens from the layers station