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.
The performance of deep neural networks is strongly influenced by the training dataset setup. In particular, when attributes having a strong correlation with the target attribute are present, the trained model can provide unintended prejudgments and show significant inference errors (i.e., the dataset bias problem). Various methods have been proposed to mitigate dataset bias, and their emphasis is on weakly correlated samples, called bias-conflicting samples. These methods are based on explicit bias labels involving human or empirical correlation metrics (e.g., training loss). However, such me…
University of Rochester（罗切斯特大学）；Kitware, Inc.（Kitware公司）
Deep image classifiers have been found to learn biases from datasets. To mitigate the biases, most previous methods require labels of protected attributes (e.g., age, skin tone) as full-supervision, which has two limitations: 1) it is infeasible when the labels are unavailable; 2) they are incapable of mitigating unknown biases -- biases that humans do not preconceive. To resolve those problems, we propose Debiasing Alternate Networks (DebiAN), which comprises two networks -- a Discoverer and a Classifier. By training in an alternate manner, the discoverer tries to find multiple unknown biases…
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Baidu Inc.（百度公司）；Institute for AI Industry Research, Tsinghua University（清华大学智能产业研究院）；Academy of Mathematics and Systems Science, Chinese Academy of Sciences（中国科学院数学与系统科学研究院）
Ensemble-based debiasing methods have been shown effective in mitigating the reliance of classifiers on specific dataset bias, by exploiting the output of a bias-only model to adjust the learning target. In this paper, we focus on the bias-only model in these ensemble-based methods, which plays an important role but has not gained much attention in the existing literature. Theoretically, we prove that the debiasing performance can be damaged by inaccurate uncertainty estimations of the bias-only model. Empirically, we show that existing bias-only models fall short in producing accurate uncerta…
Recent work has shown pre-trained language models capture social biases from the large amounts of text they are trained on. This has attracted attention to developing techniques that mitigate such biases. In this work, we perform an empirical survey of five recently proposed bias mitigation techniques: Counterfactual Data Augmentation (CDA), Dropout, Iterative Nullspace Projection, Self-Debias, and SentenceDebias. We quantify the effectiveness of each technique using three intrinsic bias benchmarks while also measuring the impact of these techniques on a model's language modeling ability, as w…
Word embeddings are the standard model for semantic and syntactic representations of words. Unfortunately, these models have been shown to exhibit undesirable word associations resulting from gender, racial, and religious biases. Existing post-processing methods for debiasing word embeddings are unable to mitigate gender bias hidden in the spatial arrangement of word vectors. In this paper, we propose RAN-Debias, a novel gender debiasing methodology which not only eliminates the bias present in a word vector but also alters the spatial distribution of its neighbouring vectors, achieving a bias…
Vision-language models can encode societal biases and stereotypes, but there are challenges to measuring and mitigating these multimodal harms due to lacking measurement robustness and feature degradation. To address these challenges, we investigate bias measures and apply ranking metrics for image-text representations. We then investigate debiasing methods and show that prepending learned embeddings to text queries that are jointly trained with adversarial debiasing and a contrastive loss reduces various bias measures with minimal degradation to the image-text representation.
Freie Universität Berlin（柏林自由大学）；University of Southern Denmark（南丹麦大学）；Universität der Bundeswehr München（慕尼黑联邦国防军大学）
Hate speech detection is a common downstream application of natural language processing (NLP) in the real world. In spite of the increasing accuracy, current data-driven approaches could easily learn biases from the imbalanced data distributions originating from humans. The deployment of biased models could further enhance the existing social biases. But unlike handling tabular data, defining and mitigating biases in text classifiers, which deal with unstructured data, are more challenging. A popular solution for improving machine learning fairness in NLP is to conduct the debiasing process wi…
LightFair: Towards an Efficient Alternative for Fair T2I Diffusion via Debiasing Pre-trained Text Encoders
State Key Laboratory of AI Safety, Institute of Computing Technology, CAS（人工智能安全国家重点实验室，计算技术研究所，中国科学院）；School of Computer Science and Tech., University of Chinese Academy of Sciences（中国科学院大学计算机科学与技术学院）；Peng Cheng Laboratory（鹏城实验室）
University of Louisville（路易斯维尔大学）；Amazon（亚马逊）；Meta
Bidirectional Transformer architectures are state-of-the-art sequential recommendation models that use a bi-directional representation capacity based on the Cloze task, a.k.a. Masked Language Modeling. The latter aims to predict randomly masked items within the sequence. Because they assume that the true interacted item is the most relevant one, an exposure bias results, where non-interacted items with low exposure propensities are assumed to be irrelevant. The most common approach to mitigating exposure bias in recommendation has been Inverse Propensity Scoring (IPS), which consists of down-w…
Deep neural networks for image-based screening and computer-aided diagnosis have achieved expert-level performance on various medical imaging modalities, including chest radiographs. Recently, several works have indicated that these state-of-the-art classifiers can be biased with respect to sensitive patient attributes, such as race or gender, leading to growing concerns about demographic disparities and discrimination resulting from algorithmic and model-based decision-making in healthcare. Fair machine learning has focused on mitigating such biases against disadvantaged or marginalised group…
University of Virginia（弗吉尼亚大学）；University of Georgia（佐治亚大学）；U.S. Army Research Laboratory（美国陆军研究实验室）
Graph Neural Networks (GNNs) have shown superior performance in analyzing attributed networks in various web-based applications such as social recommendation and web search. Nevertheless, in high-stake decision-making scenarios such as online fraud detection, there is an increasing societal concern that GNNs could make discriminatory decisions towards certain demographic groups. Despite recent explorations on fair GNNs, these works are tailored for a specific GNN model. However, myriads of GNN variants have been proposed for different applications, and it is costly to fine-tune existing debias…
Knowledge Graphs (KG) are gaining increasing attention in both academia and industry. Despite their diverse benefits, recent research have identified social and cultural biases embedded in the representations learned from KGs. Such biases can have detrimental consequences on different population and minority groups as applications of KG begin to intersect and interact with social spheres. This paper aims at identifying and mitigating such biases in Knowledge Graph (KG) embeddings. As a first step, we explore popularity bias -- the relationship between node popularity and link prediction accura…
Word embedding has become essential for natural language processing as it boosts empirical performances of various tasks. However, recent research discovers that gender bias is incorporated in neural word embeddings, and downstream tasks that rely on these biased word vectors also produce gender-biased results. While some word-embedding gender-debiasing methods have been developed, these methods mainly focus on reducing gender bias associated with gender direction and fail to reduce the gender bias presented in word embedding relations. In this paper, we design a causal and simple approach for…
Mitigating Shortcuts in Language Models with Soft Label Encoding
Stevens Institute of Technology（史蒂文斯理工学院）；Shanghai Jiao Tong University（上海交通大学）；New Jersey Institute of Technology（新泽西理工学院）；University of Georgia（佐治亚大学）
Michigan State University（密歇根州立大学）；TAL Education Group（好未来教育集团）
Dialogue systems play an increasingly important role in various aspects of our daily life. It is evident from recent research that dialogue systems trained on human conversation data are biased. In particular, they can produce responses that reflect people's gender prejudice. Many debiasing methods have been developed for various NLP tasks, such as word embedding. However, they are not directly applicable to dialogue systems because they are likely to force dialogue models to generate similar responses for different genders. This greatly degrades the diversity of the generated responses and im…
Mitigating Mismatch within Reference-based Preference Optimization
Sydney AI Centre, The University of Sydney（悉尼人工智能中心，悉尼大学）；CFAR, A*STAR（CFAR，A*STAR）；CSIRO, Data61（CSIRO，Data61）；Southeast University（东南大学）；Nanyang Technological University（南洋理工大学）
Pre-trained language models encode undesirable social biases, which are further exacerbated in downstream use. To this end, we propose MABEL (a Method for Attenuating Gender Bias using Entailment Labels), an intermediate pre-training approach for mitigating gender bias in contextualized representations. Key to our approach is the use of a contrastive learning objective on counterfactually augmented, gender-balanced entailment pairs from natural language inference (NLI) datasets. We also introduce an alignment regularizer that pulls identical entailment pairs along opposite gender directions cl…
Knowledge graph data are prevalent in real-world applications, and knowledge graph neural networks (KGNNs) are essential techniques for knowledge graph representation learning. Although KGNN effectively models the structural information from knowledge graphs, these frameworks amplify the underlying data bias that leads to discrimination towards certain groups or individuals in resulting applications. Additionally, as existing debiasing approaches mainly focus on the entity-wise bias, eliminating the multi-hop relational bias that pervasively exists in knowledge graphs remains an open question.…
Towards Trustworthy Audio Deepfake Detection: A Systematic Framework for Diagnosing and Mitigating Gender Bias
School of Computing, Wichita State University（维克托利亚州立大学计算机学院）；Institut national de la recherche scientifique (INRS-EMT)（国家科学研究机构（INRS-EMT））；INRS-UQO Mixed Research Unit on Cybersecurity（INRS-UQO网络安全混合研究单位）
University of Chinese Academy of Sciences（中国科学院大学）；Beihang University（北京航空航天大学）
The wide dissemination of fake news is increasingly threatening both individuals and society. Fake news detection aims to train a model on the past news and detect fake news of the future. Though great efforts have been made, existing fake news detection methods overlooked the unintended entity bias in the real-world data, which seriously influences models' generalization ability to future data. For example, 97\% of news pieces in 2010-2017 containing the entity `Donald Trump' are real in our data, but the percentage falls down to merely 33\% in 2018. This would lead the model trained on the f…
Mitigating the Structural Bias in Graph Adversarial Defenses
Department of Electrical Engineering, City University of Hong Kong（香港城市大学电子工程系）；School of Computer Science and Engineering, Sun Yat-sen University（中山大学计算机科学与工程学院）；School of Software Engineering, Sun Yat-sen University（中山大学软件工程学院）
