Offline Multi-Objective Optimization on Off-MOO-Bench MO-NAS (50th percentile)
4.21Average HV RankMOBO
Evaluation Results
| Method | Links | |
|---|---|---|
| MOBOEvaluation percentile=50th2026.06 | 4.21 | |
| Multiple Models + IOMEvaluation percentile=50th2026.06 | 5.66 | |
| End-to-EndEvaluation percentile=50th2026.06 | 5.94 | |
| Multi HeadEvaluation percentile=50th2026.06 | 6.23 | |
| Multiple Models + RoMAEvaluation percentile=50th2026.06 | 6.24 | |
| Multiple ModelsEvaluation percentile=50th2026.06 | 6.73 | |
| Multiple Models + COMEvaluation percentile=50th2026.06 | 6.91 | |
| End-to-End + PcGradEvaluation percentile=50th2026.06 | 7.56 | |
| Multi Head + PcGradEvaluation percentile=50th2026.06 | 7.56 | |
| DOMOOEvaluation percentile=50th2026.06 | 8.36 | |
| D(best)Evaluation percentile=50th2026.06 | 9.06 | |
| Multiple Models + ICTEvaluation percentile=50th2026.06 | 9.59 | |
| Multiple Models + Tri-MentoringEvaluation percentile=50th2026.06 | 11.1 | |
| ParetoFlowEvaluation percentile=50th2026.06 | 12.04 | |
| End-to-End + GradNormEvaluation percentile=50th2026.06 | 12.99 | |
| MOBO-qParEGOEvaluation percentile=50th2026.06 | 13.19 | |
| Multi Head + GradNormEvaluation percentile=50th2026.06 | 13.27 |