Table 2.
Mean score of different number of context size in our method.
| Agents | A | B | C | D | E | Avg |
|---|---|---|---|---|---|---|
| Ours w/context 0 | 4.4 | 2.8 | 2.6 | 1 | 1.2 | 2.4 (±1.38) |
| Ours w/context 1 | 11.6 | 0.2 | 1 | 17.2 | 85.2 | 23.04 (±35.48) |
| Ours w/context 3 | 136 | 148 | 143 | 140 | 140 | 141.4 (±4.45) |
| Ours w/context 5 | 149 | 154 | 150 | 146 | 146 | 149 (±3.32) |
Bold values indicate the setting which can produce the best results, i.e., the maximum value in that column, facilitating comparisons between the results.