Share your thoughts, 1 month free Claude Pro on usSee more

Safe Reinforcement Learning on Safety Gymnasium PointButton (C=25)

10.15Return

CSPO

Updated 1mo ago

Evaluation Results

Method	Links
CSPO 2026.06		10.15	24.4
PPO-Lag 2026.06		9.05	24.1
APPO 2026.06		9.01	25.2
IPO 2026.06		7.92	28.6
CPO 2026.06		7.02	25
C-TRPO 2026.06		6.5	28.9
EPO 2026.06		6.38	24.3
FOCOPS 2026.06		5.55	26.8
P3O 2026.06		4.16	26.3
CUP 2026.06		3.56	30.1
PCPO 2026.06		2.61	33.5
CPPOPID 2026.06		1.39	24