 initial model & dataset & Llama-2-70b-chat-hf & Mixtral-8x7B-Instruct-v0.1 & Qwen1.5-72B-Chat \\
 & \scalebox{0.9}[1]{MathGen} & \cellcolor[RGB]{240,230,140}{75.3} & \cellcolor[RGB]{240,230,140}{37.7} & \cellcolor[RGB]{240,230,140}{23.4} & \\
 & \scalebox{0.9}[1]{FgFactV} & \cellcolor[RGB]{240,230,140}{83.3} & \cellcolor[RGB]{240,230,140}{42.6} & \cellcolor[RGB]{240,230,140}{17.6} & \\
 & \scalebox{0.9}[1]{AnsCls} & \cellcolor[RGB]{240,230,140}{78.7} & \cellcolor[RGB]{240,230,140}{30.9} & \cellcolor[RGB]{240,230,140}{9.9} & \\
 & \scalebox{0.9}[1]{MathGen} & \cellcolor[RGB]{240,230,140}{72.7} & \cellcolor[RGB]{240,230,140}{45.8} & \cellcolor[RGB]{240,230,140}{39.2} & \\
 & \scalebox{0.9}[1]{FgFactV} & \cellcolor[RGB]{240,230,140}{82.5} & \cellcolor[RGB]{240,230,140}{24.8} & \cellcolor[RGB]{240,230,140}{11.1} & \\
 & \scalebox{0.9}[1]{AnsCls} & \cellcolor[RGB]{240,230,140}{46.4} & \cellcolor[RGB]{240,230,140}{20.6} & \cellcolor[RGB]{240,230,140}{2.4} & \\
