% Author: Simon-Pierre Boucher — simon-pierre.boucher@uqo.ca % generated by scripts/20_evaluate.py — do not edit \begin{tabular}{lrrrrrr} \toprule Arm & {Cost (USD)} & {Latency (s)} & {Tokens} & {Total cost (USD)} & {Mean $|e|$} & {N} \\ \midrule A (hedonic) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.191} & \num{96935} \\ A (hedonic spatial) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.208} & \num{8988} \\ B (hierarchical) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.182} & \num{96935} \\ C (catboost) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.177} & \num{96935} \\ C (gbm) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.179} & \num{96935} \\ C (gbm spatial) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.176} & \num{8988} \\ C (xgboost) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.180} & \num{96935} \\ D (avm) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.171} & \num{96935} \\ D (avm spatial) & \num{0.0000} & \num{0.0} & \num{0} & \num{0.00} & \num{0.161} & \num{8988} \\ \midrule E (single llm) & \num{0.0219} & \num{19.7} & \num{6596} & \num{1.75} & \num{0.166} & \num{80} \\ E (single llm small) & \num{0.0017} & \num{19.1} & \num{8133} & \num{0.51} & \num{0.182} & \num{300} \\ F (independent) & \num{0.2323} & \num{79.1} & \num{73947} & \num{18.58} & \num{0.159} & \num{80} \\ F (independent small) & \num{0.0188} & \num{78.5} & \num{89504} & \num{5.63} & \num{0.176} & \num{300} \\ G (debate r2) & \num{1.5217} & \num{336.2} & \num{521127} & \num{121.73} & \num{0.158} & \num{80} \\ G (debate r2 small) & \num{0.1253} & \num{358.4} & \num{667795} & \num{37.58} & \num{0.180} & \num{300} \\ H (debate skeptic) & \num{1.3501} & \num{304.8} & \num{459057} & \num{108.01} & \num{0.158} & \num{80} \\ I (placebo) & \num{1.3973} & \num{268.1} & \num{491998} & \num{111.78} & \num{0.157} & \num{80} \\ J (cost matched) & \num{0.2554} & \num{93.8} & \num{91937} & \num{20.43} & \num{0.163} & \num{80} \\ \bottomrule \end{tabular}