[Step-by-step Flow]

Why doesn’t 1up provide confidence scores?

Large language models (LLMs) struggle with confidence scoring. This is because LLMs are primarily trained to predict the next word in a sequence based on patterns in the data, not to directly estimate uncertainty or confidence.

1up uses a binary system to determine if a query can or cannot be answered. You might see other products that provide confidence scoring. In our view, these scores are misleading at best. Additionally, an AI model should never be used to grade its own performance. We believe it’s always best to let the reader decide on the quality of a 1up-generated response.

As a best practice, we recommend reviewing all responses to make sure they are accurate. Here’s what you can expect for each response:

[ placeholder ]

[ open a support ticket ]