Traditional deep neural networks typically generalize by interpolating between training examples and struggle when faced with novel combinations or configurations of inputs not seen during training.

Authors

• Person: Not available Organization: Springer
  A Comprehensive Review of Neuro-symbolic AI for Robustness ...

Sources

• A Comprehensive Review of Neuro-symbolic AI for Robustness ... link.springer.com Springer via serper

Referenced by nodes (1)

• deep neural networks concept