Robot Manipulation on LIBERO Long Horizon
97.1Average Performance ScoreOPENVLA-OFT
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| OPENVLA-OFT2026.05 | 97.1 | — | — | — | — | — | 97.6 | 98.4 | 97.9 | 94.5 | |
| LA-DIRECTStrategy=S22026.05 | 97.1 | — | — | — | — | — | 97.2 | 98.6 | 95.8 | 96.6 | |
| LA-ALIGNStrategy=S12026.05 | 97 | — | — | — | — | — | 97.4 | 98.6 | 97.2 | 94.8 | |
| LA-CONDStrategy=S32026.05 | 96.6 | — | — | — | — | — | 97 | 99.4 | 95.8 | 94.2 | |
| LA-TOKStrategy=S42026.05 | 95.5 | — | — | — | — | — | 97 | 100 | 92.2 | 92.6 | |
| UNIVLA2026.05 | 95.2 | — | — | — | — | — | 96.5 | 96.8 | 95.6 | 92 | |
| π02026.05 | 94.2 | — | — | — | — | — | 96.8 | 98.8 | 95.8 | 85.2 | |
| GR00T2026.05 | 93.9 | — | — | — | — | — | 94.4 | 97.6 | 93 | 90.6 | |
| BASELINE2026.05 | 93.1 | — | — | — | — | — | 96.6 | 97.2 | 92.8 | 85.8 | |
| MultitaskOracle=true2026.03 | 90.5 | 7.5 | — | — | 85.2 | 1.8 | — | — | — | — | |
| Sequential Fine-Tuning2026.03 | 89.8 | 6.8 | -2.4 | 0.5 | 86.6 | 3.3 | — | — | — | — | |
| Expert Replay2026.03 | 88.8 | 5.8 | -0.2 | -1.1 | 83.2 | -0.1 | — | — | — | — | |
| Dynamic Weight Expansion2026.03 | 88.4 | 5.4 | 0 | 0 | 83.4 | 0 | — | — | — | — | |
| Dark Experience Replay2026.03 | 87.6 | 4.6 | 0.7 | 0.7 | 84.7 | 1.3 | — | — | — | — | |
| SLCALR Strategy=Layered LR2026.03 | 86.9 | 3.9 | -1.3 | -0.2 | 86.1 | 2.7 | — | — | — | — | |
| Elastic Weight Consolidation2026.03 | 86.6 | 3.6 | 0.8 | 3 | 86.5 | 3.1 | — | — | — | — | |
| RETAINStrategy=Weight Merging2026.03 | 86.2 | 3.2 | 1.6 | 1 | 86.9 | 3.6 | — | — | — | — | |
| π0-FAST2026.05 | 85.5 | — | — | — | — | — | 96.4 | 96.8 | 88.6 | 60.2 | |
| THINKACT2026.05 | 84.4 | — | — | — | — | — | 88.3 | 91.4 | 87.1 | 70.9 | |
| LAPA2026.05 | 65.7 | — | — | — | — | — | 73.8 | 74.6 | 58.8 | 55.4 |