Core concepts

Logs

Logs contain the inputs and outputs of each time a Prompt, Tool or Evaluator is called.

All Prompts, Tools and Evaluators produce Logs. A Log contains the inputs and the outputs and tracks which version of Prompt/Tool/Evaluator was used.

For the example of a Prompt above, the Log would have one input called ‘topic’ and the output will be the completion.

A Log which contains an input query — A Log which contains an input called query and which resulted in two tool calls from the model.