TASK |
DATASET |
MODEL |
METRIC NAME |
METRIC VALUE |
GLOBAL RANK |
REMOVE |
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^5 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
32.02±11.72
|
# 1
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^5 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
25.52±7.92
|
# 1
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^5 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
22.31±5.17
|
# 2
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^6 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
31.51±18.69
|
# 3
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^6 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
23.35±9.48
|
# 5
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 10^6 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
20.63±1.71
|
# 4
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 2*10^6 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
31.87±18.42
|
# 4
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 2*10^6 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
21.31±6.60
|
# 7
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 2*10^6 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
18.38±6.64
|
# 5
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 5*10^5 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
26.50±13.15
|
# 3
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 5*10^5 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
24.73±8.90
|
# 5
|
|
Unsupervised Reinforcement Learning
|
URLB (pixels, 5*10^5 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
17.69±1.96
|
# 4
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^5 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
74.81±36.92
|
# 6
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^5 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
25.75±12.45
|
# 9
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^5 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
48.98±8.60
|
# 7
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^6 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
76.23±35.72
|
# 5
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^6 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
33.42±8.83
|
# 7
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 10^6 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
58.55±4.87
|
# 5
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 2*10^6 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
68.98±35.96
|
# 7
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 2*10^6 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
47.73±9.05
|
# 6
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 2*10^6 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
61.19±6.18
|
# 2
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 5*10^5 frames)
|
ProtoRL
|
Walker (mean normalized return)
|
74.72±38.85
|
# 6
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 5*10^5 frames)
|
ProtoRL
|
Quadruped (mean normalized return)
|
28.45±11.21
|
# 8
|
|
Unsupervised Reinforcement Learning
|
URLB (states, 5*10^5 frames)
|
ProtoRL
|
Jaco (mean normalized return)
|
57.35±4.31
|
# 6
|
|