Offline Reinforcement Learning on D4RL Locomotion Full datasets

102.29Hopper Score (m)ReBRAC

Evaluation results

Method
ReBRAC2026.03	102.29	94.98	109.45	—	64.04	51.18	103.8	—	85.82	84.25	111.86	—	89.74	807.67
EDAC2026.03	101.7	99.66	105.19	—	67.7	62.06	104.76	—	93.36	87.1	114.75	—	92.92	836.28
A2PR2026.01	100.8	101.5	112.1	115	68.6	56.6	98.3	103.2	89.7	94.4	114.7	114.2	97.4	—
GEM2026.03	100.08	101.23	110.33	—	54.42	49.6	100.95	—	83.51	97.02	109.74	—	89.65	806.88
DOGE2026.01	98.6	76.2	102.7	107.4	40.6	42.8	78.7	93.5	86.8	87.3	110.4	107.3	86.1	—
SSAR2026.03	95.4	101.5	106.7	—	60	51.7	98.5	—	86.4	94.1	112.4	—	89.63	806.7
BPPO2026.01	93.9	92.5	112.8	113.2	44	41	92.5	95.3	83.6	77.6	113.1	113.8	89.4	—
Diffusion-QL2026.03	90.5	101.3	111.1	—	51.1	47.8	96.8	—	87	95.5	110.1	—	87.91	791.2
TSRL2026.01	86.7	78.7	95.9	110	48.2	42.2	92	94.3	77.5	66.1	106.4	110.2	84	—
CQL+C42026.01	85.9	100.7	89.4	110.1	48.5	44.7	91.6	93.2	81.8	90.9	108.6	109.7	87.9	—
IQL2026.03	67.53	97.43	107.42	—	48.31	44.46	94.74	—	80.91	82.15	111.72	—	81.63	734.67
IQL2026.01	66.3	94.7	91.5	99.3	47.4	44	86.7	88.9	78.3	73.9	109.6	109.7	82.5	—
IDQL2026.03	65.4	92.1	108.6	—	51.1	45.9	95.9	—	82.5	85.1	112.7	—	82.14	739.3
DT2026.03	65.1	81.77	110.44	—	42.2	38.91	91.55	—	67.63	59.8	107.11	—	73.84	664.51
AWAC2026.03	63.02	98.88	101.9	—	50.02	45.13	95	—	68.52	80.62	111.44	—	79.39	714.53
TD3+BC2026.03	60.3	64.42	101.17	—	48.1	44.84	90.78	—	82.71	85.62	110.03	—	76.45	687.97
TD3BC2026.01	59.3	60.9	98	100.1	48.3	44.6	90.7	82.1	83.7	81.8	110.1	108.2	80.7	—
CQL2026.03	59.08	95.11	99.26	—	47.04	45.04	95.63	—	80.75	73.09	109.56	—	78.28	704.56
CQL2026.01	58.5	95	105.4	98.4	41	45.5	91.6	95.6	72.5	77.2	108.8	110.3	83.3	—
BC2026.01	52.9	18.1	52.5	108	42.6	55.2	55.2	92.2	75.3	26	107.5	107.9	66.1	—
SAC-N2026.03	40.82	100.33	101.31	—	68.2	60.7	98.96	—	87.47	78.99	114.93	—	83.52	751.71