Selection Bias Induced Spurious Correlations in Large Language Models

2022-07-18 · Emily McMilin

In this work we show how large language models (LLMs) can learn statistical dependencies between otherwise unconditionally independent variables due to dataset selection bias. To demonstrate the effect, we developed a masked gender task that can be applied to BERT-family models to reveal spurious correlations between predicted gender pronouns and a variety of seemingly gender-neutral variables like date and location, on pre-trained (unmodified) BERT and RoBERTa large models. Finally, we provide an online demo, inviting readers to experiment further.

📄 PDF Abstract BibTeX arXiv:2207.08982

Code (1)

2dot71mily/spurious_correlations_icml_2022 공식 구현

Tasks

Selection bias

Methods 이 논문이 사용한 방법론

Multi-Head Attention 설명 없음

Attention 설명 없음

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.

Residual Connection 설명 없음

Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…

Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…

Weight Decay 설명 없음

Linear Warmup With Linear Decay Linear Warmup With Linear Decay is a learning rate schedule in which we increase the learning rate linearly for $n$ updates and then linearly decay afterwards.

Selection Collider Bias in Large Language Models

2022-08-22 · Emily McMilin

In this paper we motivate the causal mechanisms behind sample selection induced collider bias (selection collider bias) that can cause Large Language Models (LLMs) to learn unconditional dependence between entities that …

Benchmarking Spurious Bias in Few-Shot Image Classifiers

2024-09-04 · Guangtao Zheng, Wenqian Ye, Aidong Zhang

Few-shot image classifiers are designed to recognize and classify new data with minimal supervision and limited data but often show reliance on spurious correlations between classes and spurious attributes, known as spur…

Attribute Benchmarking Few-Shot Learning

Reducing Spurious Correlations for Answer Selection by Feature Decorrelation and Language Debiasing

2022-10-01 · COLING 2022 10 · Zeyi Zhong, Min Yang, Ruifeng Xu

Deep neural models have become the mainstream in answer selection, yielding state-of-the-art performance. However, these models tend to rely on spurious correlations between prediction labels and input features, which in…

Answer Selection Contrastive Learning

The Impact of Coreset Selection on Spurious Correlations and Group Robustness

2025-07-15 · Amaya Dharmasiri, William Yang, Polina Kirichenko, Lydia Liu 외 arxiv

Coreset selection methods have shown promise in reducing the training data size while maintaining model performance for data-efficient machine learning. However, as many datasets suffer from biases that cause models to l…

Mitigating Spurious Correlations with Memorization-Guided Dataset De-Biasing

2026-06-01 · Arda Fazla, Abolfazl Hashemi arxiv

Real-world datasets often contain spurious correlations that are not causally related to the target label. When such correlations dominate the majority of training samples, models tend to rely on them, leading to misclas…