Q: 9
Universal Containers (UC) is using Einstein Generative AI to generate an account summary. UC aims
to ensure the content is safe and inclusive, utilizing the Einstein Trust Layer's toxicity scoring to
assess the
content's safety level.
What does a safety category score of 1 indicate in the Einstein Generative Toxicity Score?
Options
Discussion
Option B, Score of 1 in Einstein Trust Layer means safe content, not unsafe-it's a bit different from other vendors. I remember seeing this in the docs, but happy to be corrected if there's a recent change.
Surprised they use 1 for "safe" because in some systems 1 is lowest safety, but here Salesforce docs clearly say 1 means safe. Pick B unless they flip their rating scale in the future. Anyone see a different implementation?
B . Unless Salesforce flips their scale so higher numbers mean safer, which would change everything.
B tbh, the trap here is thinking 1 might mean "not safe" since lots of APIs do it that way. But Salesforce Trust Layer uses 1 for fully safe content. Could see people tripping on that, but docs are clear.
Call it B, but check the official guide or practice test to be sure. Safety scale directions sometimes get tricky.
C or B. A is tempting if you're used to other APIs where 1 is bad, but Salesforce calls 1 safe here.
Its B, official docs and practice tests both show that a score of 1 in the Einstein Trust Layer means safe content. Worth double-checking with the exam guide just in case I'm missing an update, but pretty sure this is it.
Why does Salesforce use 1 for safe but not zero? Seems backwards compared to some toxicity APIs.
C looks tempting since 'moderately safe' just sounds like a low-risk score, and sometimes with other APIs low numbers aren't always totally safe. I guess I'm used to seeing 1 mean 'borderline' elsewhere. Anyone else think C could fit for the Trust Layer?
Why do some APIs flag 1 as unsafe but here it's safe? Does anyone know if Salesforce ever changed this logic?
Be respectful. No spam.
