Guide-yakabatana mibvunzo · Zviri nyore Nhanho

Ruzivo Chekuchera Mazuva muLLMs Mibvunzo

Inoedza kuti ruzivo rwekucheka chii, nei mamodheru achifunga zvisirizvo zuva rawo rekugumira, uye kutsvaga, kutora uye jekiseni remazuva zvinochengeta sei mhinduro dziripo.

Nzira dzekutungamira dzinoenderanaZivo Cutoff Mazuva MuLLMs
Mubvunzo 1 ye 8

What is a model's knowledge cutoff?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. What is a model's knowledge cutoff?
    • The date the model stops working
    • The approximate date after which its training data contains little or no information
    • The date the model was released to the public
    • The date the vendor ends customer support
  2. Why are models often unsure of, or underestimate, their own cutoff?
    • Vendors deliberately hide it from the model
    • Models delete their most recent data after training
    • Models count days incorrectly
    • The final months before the cutoff are thinly represented, because people keep writing about events long afterward
  3. What is the most reliable way to find a model's knowledge cutoff?
    • Check the vendor's published model documentation
    • Ask the model directly
    • Count the words in its answers
    • Look at the model's logo
  4. What simple fix corrects many cases of a model assuming the wrong current date?
    • Fine-tuning the model every week
    • Turning the model off and on
    • Putting the current date in the system prompt
    • Using shorter prompts
  5. A chatbot with web browsing answers a question about last week's news. Which statement is accurate?
    • Its training cutoff has moved to last week
    • It can no longer make factual errors
    • It retrains itself on every query
    • Its training cutoff is unchanged; it is reading new sources and is only as accurate as the pages it retrieves
  6. Why is retrieval usually preferred over fine-tuning for keeping facts current?
    • Fine-tuning is slower and more costly as a way to keep facts current
    • Fine-tuning is impossible for large models
    • Retrieval permanently changes the model's weights
    • Retrieval works without any documents
  7. Which of these is a typical sign of stale knowledge?
    • The model refuses every question
    • The model switches to another language
    • The model names an outdated software version as the latest
    • The model's answers are unusually short
  8. When building evals for cutoff-related problems, what should you include?
    • Only questions about events well before the cutoff
    • Questions whose correct answers changed after the cutoff
    • Questions with all dates removed
    • A question asking the model to state its cutoff