IMPALA (deep)

Reported on 114 benchmarks across 2 tasks · 1 paper · 20 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Playing Games114 results

Atari GamesonAtari 2600 Space Invaders
Score· 2018-02-05
43595.78
best: 154380 (GDI-H3(200M frames))
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Asteroids
Score· 2018-02-05
108590.05
best: 760005 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Demon Attack
Score· 2018-02-05
132826.98
best: 787985 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Phoenix
Score· 2018-02-05
210996.45
best: 959580 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Alien
Score· 2018-02-05
15962.1
best: 741812.63 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Pitfall!
Score· 2018-02-05
-1.66
best: 102571 (Go-Explore)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Zaxxon
Score· 2018-02-05
32935.5
best: 725853.9 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Star Gunner
Score· 2018-02-05
200625
best: 839573.53 (Agent57)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 River Raid
Score· 2018-02-05
29608.05
best: 323417.18 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Up and Down
Score· 2018-02-05
332546.75
best: 986440 (GDI-I3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Space Invaders
Score· 2018-02-05
43595.78
best: 154380 (GDI-H3(200M frames))
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Asteroids
Score· 2018-02-05
108590.05
best: 760005 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Demon Attack
Score· 2018-02-05
132826.98
best: 787985 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Phoenix
Score· 2018-02-05
210996.45
best: 959580 (GDI-H3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Alien
Score· 2018-02-05
15962.1
best: 741812.63 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Pitfall!
Score· 2018-02-05
-1.66
best: 102571 (Go-Explore)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Zaxxon
Score· 2018-02-05
32935.5
best: 725853.9 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Star Gunner
Score· 2018-02-05
200625
best: 839573.53 (Agent57)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 River Raid
Score· 2018-02-05
29608.05
best: 323417.18 (MuZero)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Up and Down
Score· 2018-02-05
332546.75
best: 986440 (GDI-I3)
SOTA
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Boxing
Score· 2018-02-05
99.96
best: 100 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Skiing
Score· uses extra data· 2018-02-05
-10180.38
best: 0 (Best Learner)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Double Dunk
Score· 2018-02-05
-0.33
best: 24 (UCT)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Ms. Pacman
Score· 2018-02-05
7342.32
best: 243401.1 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Centipede
Score· 2018-02-05
11049.75
best: 1422628 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Tutankham
Score· 2018-02-05
292.11
best: 2354.91 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Pong
Score· 2018-02-05
20.98
best: 21 (Duel noop)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Krull
Score· 2018-02-05
8147.4
best: 594540 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Breakout
Score· 2018-02-05
787.34
best: 864 (GDI-H3(200M frames))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Frostbite
Score· 2018-02-05
317.75
best: 631378.53 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Yars Revenge
Score· 2018-02-05
84231.14
best: 998532.37 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Gopher
Score· 2018-02-05
66782.3
best: 488830 (GDI-I3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 James Bond
Score· 2018-02-05
601.5
best: 620780 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Amidar
Score· 2018-02-05
1554.79
best: 29660.08 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Tennis
Score· 2018-02-05
0.55
best: 24 (GDI-I3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Crazy Climber
Score· 2018-02-05
136950
best: 565909.85 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Gravitar
Score· 2018-02-05
359.5
best: 19213.96 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Time Pilot
Score· 2018-02-05
48481.5
best: 476763.9 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Battle Zone
Score· 2018-02-05
20885
best: 934134.88 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Beam Rider
Score· 2018-02-05
32463.47
best: 454993.53 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Asterix
Score· 2018-02-05
300732
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Kung-Fu Master
Score· 2018-02-05
43375.5
best: 1666665 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Bowling
Score· 2018-02-05
59.92
best: 260.13 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Kangaroo
Score· 2018-02-05
1632
best: 24034.16 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Assault
Score· 2018-02-05
19148.47
best: 143972.03 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Fishing Derby
Score· 2018-02-05
44.85
best: 91.16 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Seaquest
Score· 2018-02-05
1753.2
best: 1000000 (GDI-H3(200M frames))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Chopper Command
Score· 2018-02-05
28255
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Solaris
Score· 2018-02-05
2365
best: 44199.93 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Surround
Score· 2018-02-05
7.56
best: 10 (NoisyNet-Dueling)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Video Pinball
Score· 2018-02-05
572898.27
best: 999383.2 (R2D2)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Wizard of Wor
Score· 2018-02-05
9157.5
best: 197126 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Defender
Score· 2018-02-05
185203
best: 993010 (CGP)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Robotank
Score· 2018-02-05
12.96
best: 131.13 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Name This Game
Score· 2018-02-05
21537.2
best: 157177.85 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Ice Hockey
Score· 2018-02-05
3.48
best: 481.9 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Berzerk
Score· 2018-02-05
1852.7
best: 197376 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Atlantis
Score· 2018-02-05
849967.5
best: 3837300 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 HERO
Score· 2018-02-05
33730.55
best: 114736.26 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Bank Heist
Score· 2018-02-05
1223.15
best: 27219.8 (MuZero (Res2 Adam))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Private Eye
Score· 2018-02-05
98.5
best: 95756 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Q*Bert
Score· 2018-02-05
351200.12
best: 580328.14 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Road Runner
Score· 2018-02-05
57121
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Boxing
Score· 2018-02-05
99.96
best: 100 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Skiing
Score· uses extra data· 2018-02-05
-10180.38
best: 0 (Best Learner)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Double Dunk
Score· 2018-02-05
-0.33
best: 24 (UCT)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Ms. Pacman
Score· 2018-02-05
7342.32
best: 243401.1 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Centipede
Score· 2018-02-05
11049.75
best: 1422628 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Tutankham
Score· 2018-02-05
292.11
best: 2354.91 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Pong
Score· 2018-02-05
20.98
best: 21 (Duel noop)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Krull
Score· 2018-02-05
8147.4
best: 594540 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Breakout
Score· 2018-02-05
787.34
best: 864 (GDI-H3(200M frames))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Frostbite
Score· 2018-02-05
317.75
best: 631378.53 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Yars Revenge
Score· 2018-02-05
84231.14
best: 998532.37 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Gopher
Score· 2018-02-05
66782.3
best: 488830 (GDI-I3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 James Bond
Score· 2018-02-05
601.5
best: 620780 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Amidar
Score· 2018-02-05
1554.79
best: 29660.08 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Tennis
Score· 2018-02-05
0.55
best: 24 (GDI-I3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Crazy Climber
Score· 2018-02-05
136950
best: 565909.85 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Gravitar
Score· 2018-02-05
359.5
best: 19213.96 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Time Pilot
Score· 2018-02-05
48481.5
best: 476763.9 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Battle Zone
Score· 2018-02-05
20885
best: 934134.88 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Beam Rider
Score· 2018-02-05
32463.47
best: 454993.53 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Asterix
Score· 2018-02-05
300732
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Kung-Fu Master
Score· 2018-02-05
43375.5
best: 1666665 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Bowling
Score· 2018-02-05
59.92
best: 260.13 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Kangaroo
Score· 2018-02-05
1632
best: 24034.16 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Assault
Score· 2018-02-05
19148.47
best: 143972.03 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Fishing Derby
Score· 2018-02-05
44.85
best: 91.16 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Seaquest
Score· 2018-02-05
1753.2
best: 1000000 (GDI-H3(200M frames))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Chopper Command
Score· 2018-02-05
28255
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Solaris
Score· 2018-02-05
2365
best: 44199.93 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Surround
Score· 2018-02-05
7.56
best: 10 (NoisyNet-Dueling)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Video Pinball
Score· 2018-02-05
572898.27
best: 999383.2 (R2D2)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Wizard of Wor
Score· 2018-02-05
9157.5
best: 197126 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Defender
Score· 2018-02-05
185203
best: 993010 (CGP)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Robotank
Score· 2018-02-05
12.96
best: 131.13 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Name This Game
Score· 2018-02-05
21537.2
best: 157177.85 (MuZero)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Ice Hockey
Score· 2018-02-05
3.48
best: 481.9 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Berzerk
Score· 2018-02-05
1852.7
best: 197376 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Atlantis
Score· 2018-02-05
849967.5
best: 3837300 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 HERO
Score· 2018-02-05
33730.55
best: 114736.26 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Bank Heist
Score· 2018-02-05
1223.15
best: 27219.8 (MuZero (Res2 Adam))
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Private Eye
Score· 2018-02-05
98.5
best: 95756 (Go-Explore)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Q*Bert
Score· 2018-02-05
351200.12
best: 580328.14 (Agent57)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Video GamesonAtari 2600 Road Runner
Score· 2018-02-05
57121
best: 999999 (GDI-H3)
IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures arXiv:1802.01561
Atari GamesonAtari 2600 Freeway
Score
0
best: 34 (TRPO-hash)
Atari GamesonAtari 2600 Enduro
Score
0
best: 14330 (GDI-I3)
Atari GamesonAtari 2600 Montezuma's Revenge
Score
0
best: 43791 (Go-Explore)
Atari GamesonAtari 2600 Venture
Score
0
best: 2623.71 (Agent57)
Video GamesonAtari 2600 Freeway
Score
0
best: 34 (TRPO-hash)
Video GamesonAtari 2600 Enduro
Score
0
best: 14330 (GDI-I3)
Video GamesonAtari 2600 Montezuma's Revenge
Score
0
best: 43791 (Go-Explore)
Video GamesonAtari 2600 Venture
Score
0
best: 2623.71 (Agent57)