How will you confirm the authentication of ML models?
There will be an intelligent verification algorithm, and also data and ml models validators.
tell me about the algorithm more

The algorithm will check for the similarity of neural networks to each other so that the platform does not have networks with same tasks, as well as checking for white noise and validity of some data.
It's interesting. Can you explain what's the definition of the similarity in your algorithm? How the similarity is measured in what metric/space ?
The similarity of models is determined by many criteria: one of them is the structure of layers, a description of the metadata of the model and which task is solved by the model.