AINeutralarXiv – CS AI · 8h ago6/10
🧠
Paradoxical noise preference in RNNs
Researchers discovered that continuous-time RNNs trained with noise injected inside activation functions paradoxically perform best when noise remains present at test time, contradicting conventional assumptions about noise removal. This phenomenon stems from noise-induced shifts in neural network dynamics that become computationally integrated into learned representations, revealing that networks can overfit to training noise itself rather than just input-output mappings.