About the dormant threshold setting in the ReDO paper

[ZifanWu]

Hello there,

There seems to be an inconsistency on the default dormant threshold setting ($\tau$ in the paper) between the main paper and the appendix:

In the main paper (Page 5, Section 5), it says:

For agents trained with ReDo, we use a threshold of τ = 0.1, unless otherwise noted, as we found this gave a better performance than using a threshold of 0 or 0.025.

In the appendix (Table 1), it says:

0.025 for default setting, 0.1 otherwise

Could you please specify the best setting for this hyperparameter? Thanks!

[GhadaSokar]

Hello,

Thank you for your question.

We used τ = 0.1 for all settings except in the default where we use τ = 0.025.

[ZifanWu]

Understood. Thank you for the reply. If I understand correctly, the default setting (0.025) is for demonstrating the dormant phenomenon and 0.1, under which the perf is better, is for evaluation?