Multi-agent coordination on FindGoal (test)
103.5Average Episode Lengthae-comm
Evaluation results
| Method | Links | |
|---|---|---|
| ae-commtraining_method=autoencoded communication2021.10 | 103.5 | |
| rl-comm w/ biastraining_method=reinforcement learning, inductive_biases=positive signaling and listening2021.10 | 158 | |
| no-commcommunication=none2021.10 | 169 | |
| rl-commtraining_method=reinforcement learning2021.10 | 184.8 |