verifier                   | final true success   | seeds | true, steps 30-40    | final verifier reward
(a) flip eps=0.0           | [0.999 0.987 0.98 ]  | 8/8 | [0.964 0.669 0.348]  | [0.999 0.987 0.98 ]
(a) flip eps=0.1           | [0.998 0.986 0.959]  | 8/8 | [0.92  0.564 0.212]  | [0.895 0.909 0.872]
(a) flip eps=0.25          | [0.987 0.877 0.687]  | 0/8 | [0.853 0.397 0.092]  | [0.734 0.674 0.586]
(a) flip eps=0.4           | [0.83  0.427 0.098]  | 0/8 | [0.761 0.255 0.05 ]  | [0.56  0.497 0.433]
(b) false-pos q=0.5        | [0.98  0.808 0.518]  | 2/8 | [0.861 0.302 0.094]  | [0.988 0.884 0.726]
(b) false-pos q=1.0        | [0.486 0.006 0.001]  | 0/8 | [0.523 0.031 0.006]  | [1.    0.996 0.999]
(c) binary                 | [0.999 0.987 0.98 ]  | 8/8 | [0.964 0.669 0.348]  | [0.999 0.987 0.98 ]
(c) partial count/T        | [1.    1.    0.992]  | 8/8 | [0.983 0.789 0.47 ]  | [0.973 0.947 0.986]
(c) binary + 0.25 bonus    | [1.    0.983 0.98 ]  | 8/8 | [0.964 0.688 0.353]  | [1.25  1.232 1.23 ]
