References of success of the WTTE-RNN structure?

[SoufianeDataFan]

I am frequent users of the WTTE-RNN and believe the idea of this structure fits very well, theoretically, the problems of survival analysis and Churn predictions.

I have applied this model on medical data to predict sepsis but it doesn't seem to be working well compared to the basic light-GBM applied on time series.

I am curious to know if someone is aware of some papers that talking about this model more successfully.

I am trying to be negative, but as far as I know: WTTE RNN sounds good but doesn't work :(

Correct me please if I am wrong!