Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Continual Reinforcement Learning on Atari Two-cycle (train)
Loading...
0.722
C1 Forward Score
DV3
-0.06632
0.13834
0.343
0.54766
Mar 12, 2026
C1 Forward Score
C2 Forward Score
Max Forward Score
C1 Forward Transfer
C2 Forward Transfer
Recovery Score
Accuracy
Minimum Accuracy
Worst-Case Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
C1 Forward Score
C2 Forward Score
Max Forward Score
C1 Forward Transfer
C2 Forward Transfer
Recovery Score
Accuracy
Minimum Accuracy
Worst-Case Accuracy
DV3
Training Mode=Two-cycle
2026.03
0.722
0.378
0.735
-0.514
-0.75
0.61
0.9
-0.393
-0.299
TES-SAC
Training Mode=Two-cycle
2026.03
0.194
0.112
0.089
-0.898
-0.882
0.767
4.4
-0.203
-0.168
ARROW
Training Mode=Two-cycle
2026.03
-0.036
0.03
0.012
-0.554
0.309
1.418
79.6
0.442
0.388
Feedback
Search any
task
Search any
task
