Complete result settings for 20 Newsgroup dataset (table columns are sortable)
The experiments are described in Rooshenas and Lowd, Discriminative Structure Learning of Arithmetic Circuits, AIStats 16

Standard Deviation L1 Penalty Train CLL Validation CLL Test CLL Node# Edge# Feature# Learning Time
0.1 0.1 -70.222068 -62.588482 -77.889253 92615 180971 9358 37808.112292
0.5 0.1 -70.222070 -62.588424 -77.889236 92615 180971 9358 39854.436203
0.1 0.5 -70.216332 -62.584225 -77.876932 92615 180971 9358 41056.562453
0.5 1 -70.195543 -62.571863 -77.871379 92615 180971 9358 44876.476738
0.5 0.5 -70.207251 -62.578385 -77.876717 92615 180971 9358 45118.807897
0.1 1 -70.197237 -62.571491 -77.871898 92615 180971 9358 45714.698309
0.1 2 -70.180530 -62.560402 -77.831158 92615 180971 9358 60117.698720
0.5 2 -70.180336 -62.561901 -77.830930 92615 180971 9358 64384.729033