# relevantfields Use `relevantfields` to automatically identify which fields in your log data are most relevant to a given condition. The command compares the baseline logs against the subset matching your condition, and returns fields ranked by their relevance score. **Syntax** ``` relevantfields [{{field1}}, {{field2}}, ...] where {{condition}} ``` The field list is optional. If omitted, all fields are analyzed. The command returns the following output fields: + `@fieldName` — Name of the field + `@relevanceScore` — A score indicating relevancy on a scale of 0 to 1 + `@topRelevanceContributors` — For categorical fields, shows entries with the highest shifts in frequency. + `@conditionalMedian` — For numerical fields, the median of values in logs that meet the condition + `@baselineMedian` — For numerical fields, the median of values in logs that do not meet the condition **Examples** **Analyze all fields for slow API responses** ``` relevantfields where Time > 400 ``` **Focus analysis on specific fields** ``` relevantfields Controller, Path, Service where Time > 400 ``` **Identify drivers of Lambda timeouts** ``` relevantfields where Duration > 5000 ```