HardnessofNoise-FreeLearningfor Two-Hidden-LayerNeuralNetworks
–Neural Information Processing Systems
We give superpolynomial statistical query (SQ) lower bounds for learning twohidden-layer ReLU networks with respect to Gaussian inputs in the standard (noise-free)model.
Neural Information Processing Systems
Feb-8-2026, 15:56:30 GMT
- Country:
- Africa (0.04)
- Asia > Middle East
- UAE > Abu Dhabi Emirate > Abu Dhabi (0.04)
- Genre:
- Research Report (0.46)
- Technology: