% Key papers for Hardness of learning neural networks (scime atlas, https://atlas.scime.ai). Authors are not included: add them from the paper pages.

@inproceedings{aaai2017_when_104536,
  title = {{When and Why Are Deep Networks Better Than Shallow Ones?}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2017},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/10913},
}

@inproceedings{iclr2017_why_28327,
  title = {{Why Deep Neural Networks for Function Approximation?}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2017},
  url = {https://openreview.net/forum?id=SkpSlKIel},
}

@inproceedings{aaai2016_on_105071,
  title = {{On the Depth of Deep Neural Networks: A Theoretical View}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2016},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/10243},
}

@inproceedings{neurips2017_approximation_56162,
  title = {{Approximation and Convergence Properties of Generative Adversarial Learning}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2017},
  url = {https://nips.cc/virtual/2017/poster/9329},
}

@inproceedings{neurips2019_efficient_58184,
  title = {{Efficient Approximation of Deep ReLU Networks for Functions on Low Dimensional Manifolds}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2019},
  url = {https://nips.cc/virtual/2019/poster/13887},
}

@inproceedings{neurips2018_deep_57005,
  title = {{Deep Homogeneous Mixture Models: Representation, Separation, and Approximation}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2018},
  url = {https://nips.cc/virtual/2018/poster/11687},
}

@inproceedings{neurips2017_f_56781,
  title = {{f-GANs in an Information Geometric Nutshell}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2017},
  url = {https://nips.cc/virtual/2017/poster/8842},
}

@inproceedings{neurips2020_hardness_59992,
  title = {{Hardness of Learning Neural Networks with Natural Weights}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2020},
  url = {https://nips.cc/virtual/2020/poster/18653},
}

@inproceedings{neurips2020_logarithmic_60267,
  title = {{Logarithmic Pruning is All You Need}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2020},
  url = {https://nips.cc/virtual/2020/poster/18706},
}

@inproceedings{iclr2020_on_29531,
  title = {{On Universal Equivariant Set Networks}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2020},
  url = {https://openreview.net/forum?id=HkxTwkrKDB},
}

@inproceedings{neurips2021_on_63319,
  title = {{On the Representation Power of Set Pooling Networks}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2021},
  url = {https://nips.cc/virtual/2021/poster/26283},
}

@inproceedings{icml2022_synergy_41230,
  title = {{Synergy and Symmetry in Deep Learning: Interactions between the Data, Model, and Inference Algorithm}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2022},
  url = {https://icml.cc/virtual/2022/poster/16697},
}

@inproceedings{neurips2024_hardness_73108,
  title = {{Hardness of Learning Neural Networks under the Manifold Hypothesis}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2024},
  url = {https://neurips.cc/virtual/2024/poster/94321},
}

@inproceedings{neurips2020_the_60947,
  title = {{The Implications of Local Correlation on Learning Some Deep Functions}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2020},
  url = {https://nips.cc/virtual/2020/poster/17651},
}

@inproceedings{neurips2023_optimal_69815,
  title = {{Optimal approximation using complex-valued neural networks}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2023},
  url = {https://nips.cc/virtual/2023/poster/70134},
}

@inproceedings{icml2024_vocabulary_44343,
  title = {{Vocabulary for Universal Approximation: A Linguistic Perspective of Mapping Compositions}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/34109},
}

@inproceedings{aaai2025_on_10177,
  title = {{On the Hardness of Training Deep Neural Networks Discretely}},
  booktitle = {AAAI Conference on Artificial Intelligence (AAAI)},
  year = {2025},
  url = {https://ojs.aaai.org/index.php/AAAI/article/view/33797},
}

@inproceedings{icml2022_deep_40421,
  title = {{Deep Network Approximation in Terms of Intrinsic Parameters}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2022},
  url = {https://icml.cc/virtual/2022/poster/18197},
}

@inproceedings{neurips2022_effects_64894,
  title = {{Effects of Data Geometry in Early Deep Learning}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2022},
  url = {https://nips.cc/virtual/2022/poster/53764},
}

@inproceedings{neurips2023_computational_66867,
  title = {{Computational Complexity of Learning Neural Networks: Smoothness and Degeneracy}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2023},
  url = {https://nips.cc/virtual/2023/poster/72775},
}

@inproceedings{icml2024_benign_43885,
  title = {{Benign Overfitting in Two-Layer ReLU Convolutional Neural Networks for XOR Data}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/34567},
}

@inproceedings{neurips2025_on_74699,
  title = {{On the Hardness of Approximating Distributions with Tractable Probabilistic Models}},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  year = {2025},
  url = {https://openreview.net/forum?id=0GvEaa9prl},
}

@inproceedings{iclr2026_improving_85900,
  title = {{Improving Set Function Approximation with Quasi-Arithmetic Neural Networks}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2026},
  url = {https://openreview.net/forum?id=JQR8OCptcG},
}

@inproceedings{icml2024_characterizing_45809,
  title = {{Characterizing ResNet's Universal Approximation Capability}},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2024},
  url = {https://icml.cc/virtual/2024/poster/32643},
}

@inproceedings{iclr2026_tractability_85163,
  title = {{Tractability via Low Dimensionality: The Parameterized Complexity of Training Quantized Neural Networks}},
  booktitle = {International Conference on Learning Representations (ICLR)},
  year = {2026},
  url = {https://openreview.net/forum?id=BAQNrsr987},
}