Offline Reinforcement Learning on halfcheetah medium-replay v2
51.2Normalized ScoreReBRAC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ReBRACCategory=Non-diffusion, Delay=No delay2026.03 | 51.2 | — | — | |
| SAIDCategory=Diffusion Planners, Delay=02026.03 | 50.2 | — | — | |
| DQL*Category=Diffusion Policies, Delay=No delay2026.03 | 48.9 | — | — | |
| DQLCategory=Diffusion Policies, Delay=No delay2026.03 | 47.8 | — | — | |
| CQLCategory=Non-diffusion, Delay=No delay2026.03 | 46.2 | — | — | |
| DVCategory=Diffusion Planners, Delay=No delay2026.03 | 45.8 | — | — | |
| BCCategory=Non-diffusion, Delay=No delay2026.03 | 38.4 | — | — | |
| DiffuserCategory=Diffusion Planners, Delay=No delay2026.03 | 37.7 | — | — | |
| SAIDCategory=Diffusion Planners, Delay=42026.03 | 34.3 | — | — | |
| SAIDCategory=Diffusion Planners, Delay=82026.03 | 8.3 | — | — | |
| SAIDCategory=Diffusion Planners, Delay=162026.03 | 2.1 | — | — | |
| DQL+SACategory=Baseline w/ delay, Delay=82026.03 | 1.4 | — | — | |
| DQL+SACategory=Baseline w/ delay, Delay=162026.03 | 1.3 | — | — | |
| DQL+SACategory=Baseline w/ delay, Delay=42026.03 | 0.2 | — | — | |
| LRTMode=lrt2025.10 | — | 558 | 1.14 | |
| LRT+QMode=lrt+q2025.10 | — | 615 | 13.93 | |
| QMode=q2025.10 | — | 598 | 13.13 |