Clarification on ATE estimates in CausalBert.py outputs

[tomatoboy-hub]

Hello,

I tried running the experiments and got the following scores by following this:
ATE_unadjusted: -0.2240035541871202
ATE_adjusted: -0.16479750378762317
However, the outputs from CausalBert.py seem to be closer to the ATE_unadjusted values (e.g., around -0.23 or -0.22).

I was under the impression that this model is designed to estimate ATE_adjusted, but based on these results, I am wondering if this is correct. Could you clarify whether CausalBert.py is intended to produce values closer to ATE_adjusted or if it aligns more with ATE_unadjusted?

Thank you for your time and for providing this great tool!