Multi-Task Learning on SARCOS standard (test) benchmark leaderboard · SOTA2 Research

Multi-Task Learning on SARCOS standard (test)

9.92Average ErrorUW + MAdam

Evaluation Results

Method	Links
UW + MAdam2026.06		9.92	1.33	13.19	0.59	0.17	46.14	7.73	0.29
LS + MAdam2026.06		10	1.46	13.55	0.61	0.19	45.43	8.47	0.31
DWA + MAdam2026.06		10.09	1.47	13.57	0.62	0.19	45.94	8.51	0.3
CAGrad + MAdam2026.06		10.38	1.11	13.79	0.44	0.11	47.92	9.13	0.17
UW2026.06		10.83	1.45	14.29	0.61	0.18	50.71	8.27	0.3
CAGrad2026.06		10.98	1.16	14.88	0.66	0.15	50.3	9.53	0.18
LS2026.06		10.99	1.63	14.87	0.65	0.19	50.32	8.96	0.3
IMTL + MAdam2026.06		11.05	0.81	13.32	0.34	0.08	55.64	7.02	0.15
DWA2026.06		11.11	1.64	14.85	0.66	0.19	51.15	8.98	0.31
IMTL2026.06		12.1	0.84	14.86	0.39	0.08	60.69	7.7	0.15