Offline Reinforcement Learning on D4RL AntMaze v2 (all configurations) benchmark leaderboard · SOTA2 Research

Offline Reinforcement Learning on D4RL AntMaze v2 (all configurations)

100UMaze Success RateGTP

Evaluation Results

Method	Links
GTP2025.10		100	81.9	83.3	94.2	53.5	71	—	80.6
ARQ2025.10		97	62	80	82	37	58	—	69.3
QGPO2025.10		96.4	74.4	83.6	83.8	66.6	64.8	—	78.3
IDQL-A2025.10		94	80.2	84.2	84.8	63.5	67.9	—	79.1
χ-QL2025.10		93.8	82	76	73.6	46.5	49	—	70.1
D-QL2025.10		93.4	66.2	76.6	78.6	46.4	56.6	—	69.6
BDM2025.10		93	81	79	84	—	—	—	—
SCAS2024.10		90.4	63.8	76.6	80.4	49	50.6	410.8	—
IQL2024.10		89.6	65.6	76.4	72.8	42	46	392.4	—
IQL2025.10		87.5	62.2	71.2	70	39.6	47.5	—	63
OSR2024.10		87.4	55.6	22.6	19.6	0	0	185.2	—
CQL2024.10		82.6	10.2	59	46.6	16.4	3.2	218	—
SDC2024.10		81.4	49.6	55	56.6	20.8	25.8	289.2	—
DOGE2026.01		78.9	43.8	0	0	0	0	—	20.5
C-AC2025.10		75.8	77.6	56.8	—	—	—	—	—
TSRL2026.01		74.3	57.6	0	0	0	0	—	22
CQL2025.10		74	84	61.2	53.7	15.8	14.9	—	50.6
TD3BC2024.10		73	47	0	0.2	0	0	120.2	—
IQL2026.01		69.8	58.1	0.4	0	0	0	—	21.4
A2PR2026.01		69.2	27.5	0.3	0	0.2	0	—	16.2
C42026.01		68.7	65.9	2.9	7.2	7.3	9.8	—	27
BC2024.10		66.8	56.8	0	0	0	0	123.6	—
SORL2026.01		65.3	39.1	0	0	0	0	—	17.4
OneStep2024.10		54	57.8	0	0.6	0	0.2	112.6	—
BPPO2026.01		53.3	45.1	0	0.3	0	0	—	16.5
LN2026.01		48.1	29.7	0	0	0	0	—	13
DR32026.01		41.1	0.5	0	0	0	0	—	6.9
CQL2026.01		21.1	43.3	0	0	0	0	—	10.7
TD3BC2026.01		20.9	16.8	0	0	0	0	—	6.3
MOPO2024.10		0	0	0	0	0	0	0	—