# Latent Adversarial Debiasing: Mitigating Collider Bias in Deep Neural Networks

arXiv: https://arxiv.org/abs/2011.11486 · HTML: https://mlsanity.com/paper/2011.11486/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Nov 2020.

## Built on by (most cited first, 16 of 16)

- [Learning Debiased Representation via Disentangled Feature Augmentation](https://mlsanity.com/paper/2107.01372/) (2021)
- [Debiasing Graph Neural Networks via Learning Disentangled Causal Substructure](https://mlsanity.com/paper/2209.14107/) (2022)
- [BiaSwap: Removing dataset bias with bias-tailored swapping augmentation](https://mlsanity.com/paper/2108.10008/) (2021)
- [Learning Debiased Classifier with Biased Committee](https://mlsanity.com/paper/2206.10843/) (2022)
- [Group Robust Classification Without Any Group Information](https://mlsanity.com/paper/2310.18555/) (2023)
- [Mitigating Dataset Bias by Using Per-sample Gradient](https://mlsanity.com/paper/2205.15704/) (2022)
- [Revisiting the Importance of Amplifying Bias for Debiasing](https://mlsanity.com/paper/2205.14594/) (2022)
- [Bias-Conflict Sample Synthesis and Adversarial Removal Debias Strategy for Temporal Sentence Grounding in Video](https://mlsanity.com/paper/2401.07567/) (2024)
- [Improving Robustness to Multiple Spurious Correlations by Multi-Objective Optimization](https://mlsanity.com/paper/2409.03303/) (2024)
- [Enhancing Intrinsic Features for Debiasing via Investigating Class-Discerning Common Attributes in Bias-Contrastive Pair](https://mlsanity.com/paper/2404.19250/) (2024)
- [Signal Is Harder To Learn Than Bias: Debiasing with Focal Loss](https://mlsanity.com/paper/2305.19671/) (2023)
- [Hybrid Sample Synthesis-based Debiasing of Classifier in Limited Data Setting](https://mlsanity.com/paper/2312.08288/) (2023)
- [Denoising after Entropy-based Debiasing A Robust Training Method for Dataset Bias with Noisy Labels](https://mlsanity.com/paper/2212.01189/) (2022)
- [Harnessing Diffusion-Generated Synthetic Images for Fair Image Classification](https://mlsanity.com/paper/2511.08711/) (2025)
- [Learning Decomposable and Debiased Representations via Attribute-Centric Information Bottlenecks](https://mlsanity.com/paper/2403.14140/) (2024)
- [DISCO: Mitigating Bias in Deep Learning with Conditional Distance Correlation](https://mlsanity.com/paper/2506.11653/) (2025)

## Abstract (arXiv)

Collider bias is a harmful form of sample selection bias that neural networks are ill-equipped to handle. This bias manifests itself when the underlying causal signal is strongly correlated with other confounding signals due to the training data collection procedure. In the situation where the confounding signal is easy-to-learn, deep neural networks will latch onto this and the resulting model will generalise poorly to in-the-wild test scenarios. We argue herein that the cause of failure is a combination of the deep structure of neural networks and the greedy gradient-driven learning process used - one that prefers easy-to-compute signals when available. We show it is possible to mitigate against this by generating bias-decoupled training data using latent adversarial debiasing (LAD), even when the confounding signal is present in 100% of the training data. By training neural networks on these adversarial examples,we can improve their generalisation in collider bias settings. Experiments show state-of-the-art performance of LAD in label-free debiasing with gains of 76.12% on background coloured MNIST, 35.47% on fore-ground coloured MNIST, and 8.27% on corrupted CIFAR-10.