L=20 H=22; uniform-policy success per goal: [2.032e-01 1.904e-01 1.672e-01 1.375e-01 1.057e-01 7.600e-02 5.100e-02
 3.180e-02 1.850e-02 9.900e-03 4.900e-03 2.200e-03 9.000e-04 4.000e-04
 1.000e-04 0.000e+00 0.000e+00 0.000e+00 0.000e+00 0.000e+00]
mode           success@100 success@300  near (g<14) @300  far (g>=14) @300  mean ESS frac
none                 0.620       0.703             0.983             0.051            nan
naive                0.507       0.644             0.537             0.894            nan
iw                   0.527       0.553             0.463             0.762          0.248
iw_selfnorm          0.568       0.627             0.531             0.852          0.234
