https://en.wikipedia.org/wiki/Model_collapse https://en.wikipedia.org/wiki/Model_collapse - you want to use sigmoid 1.0, but the closer you are to 1.0 the highe

https://en.wikipedia.org/wiki/Model_collapse https://en.wikipedia.org/wiki/Model_collapse - you want to use sigmoid 1.0, but the closer you are to 1.0 the higher the chance your model will collapse so you use 0.99-0.98, but those lead to data loss so after n passes all the original data becomes lost so you have a strict data limit there. The rest is just the general reality I am sure you are familiar with: - https://en.wikipedia.org/wiki/Catastrophic_interference https://en.wikipedia.org/wiki/Catastrophic_interference - https://en.wikipedia.org/wiki/Fine-tuning_(deep_learning) https://en.wikipedia.org/wiki/Fine-tuning_(deep_learning) - https://en.wikipedia.org/wiki/Entropy_(information_theory) https://en.wikipedia.org/wiki/Entropy_(information_theory)