Inverse Transition Learning on Gridworld 40% stochastic-policy states benchmark leaderboard · SOTA2 Research

Inverse Transition Learning on Gridworld 40% stochastic-policy states

0.37Epsilon Matching ErrorMLE

Evaluation Results

Method	Links
MLE2024.11		0.37	0.39	0.31	0.34	0.01	141.37	-5.34	-5.18	-4.71	23.13
PS2024.11		0.38	0.39	0.29	0.31	8.65	160.05	-4.33	-3.49	-2.35	16.37
MCE2024.11		0.65	0.5	0.55	0.45	117.47	137.62	56.05	76.7	108.76	3.06
ITL2024.11		0.65	0.51	0.56	0.45	0.32	137.05	103.87	106.46	109.7	0
BITL2024.11		0.78	0.64	0.64	0.54	14,384.26	159.36	104.64	107.36	109.04	0