Robot Assembly on COHERENT Benchmark
14Mono-type Task Assembly ScorePrimitive MCTS
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Primitive MCTS2025.12 | 14 | 0 | 0 | 21.5 | 0 | 26.9 | 0 | 21.7 | |
| DMRS-2D2025.12 | 11.5 | 50 | 26.7 | 19.9 | 40 | 24.5 | 37.5 | 19.6 | |
| DMRS-1D2025.12 | 10.6 | 70 | 46.7 | 18 | 66.7 | 20.7 | 60 | 17.2 | |
| LLM-MCTS2025.12 | 10.2 | 70 | 6.7 | 20.9 | 0 | 26.9 | 20 | 20.5 | |
| CMRS2025.12 | 7.9 | 90 | 53.3 | 16.4 | 53.3 | 22.2 | 62.5 | 16.5 | |
| COHERENT2025.12 | 7.4 | 90 | 100 | 11.9 | 100 | 16.1 | 97.5 | 12.4 | |
| CLiMRS2025.12 | 6.8 | 90 | 100 | 11.5 | 100 | 13.1 | 97.5 | 10.9 | |
| Ground Truth (GT)2025.12 | 6.5 | — | — | 10.3 | — | 12.9 | — | 10.3 |