In evaluation section it is stated that
"Experiments (including ablation studies, tests on alternative architectures, and additional analysis) must be separately reported. "
Thus, if we tried 5 different architectures, and our final submitted model is based on one of them, do we need to include the results we obtained with other 4 architectures, for comparison reasons, in the 3-pages report we write?
I've a question connected to my previous thread, but similar to this thread so I'll write here.
If a team submits 8 different models in the final phase (2 per team), should the team describe all the models in the report? Furthermore, should the team provide the code to obtain all the different models in the notebook?