Controls how the agent evaluates the final result.
"llm": cross-compares the final answer to the sources, filters those sources to only the
ones used in the answer, and populates missingInformation and isPartialAnswer.
"none": the result is not evaluated and sources are not filtered.
Controls how the agent evaluates the final result.
"llm": cross-compares the final answer to the sources, filters those sources to only the ones used in the answer, and populatesmissingInformationandisPartialAnswer."none": the result is not evaluated and sources are not filtered.