```xml <result> <column_name>Lower Confidence Interval</column_name> <family>additive</family> <subtype>intensive_additive</subtype> <confidence>0.92</confidence> <requires_user_confirmation>false</requires_user_confirmation> <sum_meaningful>false</sum_meaningful> <signed>false</signed> <reducers> <primary_reducer>min</primary_reducer> <valid_reducers>min, max, mean, median</valid_reducers> <excluded_reducers>sum (summing confidence bounds is meaningless), product (not a ratio)</excluded_reducers> <average_semantics>arithmetic_mean</average_semantics> </reducers> <inner_outer_measure> <inner_measure_reducer>min</inner_measure_reducer> <outer_measure_reducer>max</outer_measure_reducer> </inner_outer_measure> <multiplicative_partner_derivable>false</multiplicative_partner_derivable> <reasoning> Table-type gate: DIMENSION with no sequence axis makes multiplicative highly implausible. The column represents a statistical bound (lower confidence interval), which is an absolute level/intensity that cannot be meaningfully summed across entities. Center-of-mass (mean=5.28, median=5.24, range 2.7-7.5) is far from both 0.0 and 1.0, confirming an absolute level rather than a ratio. No naming tokens suggest multiplicative behavior. This is intensive_additive: when volumes merge, the appropriate reducer for a lower bound is min (the most conservative lower bound), not sum or mean. </reasoning> <table_type_conflict>false</table_type_conflict> </result> ```