André F. Cruz

André F. Cruz, Moritz Hardt, Celestine Mendler-Dünner

Current question-answering benchmarks predominantly focus on accuracy in realizable prediction tasks. Conditioned on a question and answer-key, does the most likely token match the ground truth? Such benchmarks necessarily fail to evaluate LLMs' ability to quantify ground-truth outcome uncertainty. In this work, we focus on the use of LLMs as risk scores for unrealizable prediction tasks. We introduce folktexts, a software package to systematically generate risk scores using LLMs, and evaluate them against US Census data products. A flexible API enables the use of different prompting schemes, local or web-hosted models, and diverse census columns that can be used to compose custom prediction tasks. We evaluate 17 recent LLMs across five proposed benchmark tasks. We find that zero-shot risk scores produced by multiple-choice question-answering have high predictive signal but are widely miscalibrated. Base models consistently overestimate outcome uncertainty, while instruction-tuned models underestimate uncertainty and produce over-confident risk scores. In fact, instruction-tuning polarizes answer distribution regardless of true underlying data uncertainty. This reveals a general inability of instruction-tuned LLMs to express data uncertainty using multiple-choice answers. A separate experiment using verbalized chat-style risk queries yields substantially improved calibration across instruction-tuned models. These differences in ability to quantify data uncertainty cannot be revealed in realizable settings, and highlight a blind-spot in the current evaluation ecosystem that folktexts covers.

André F. Cruz, Moritz Hardt

Seven years ago, researchers proposed a postprocessing method to equalize the error rates of a model across different demographic groups. The work launched hundreds of papers purporting to improve over the postprocessing baseline. We empirically evaluate these claims through thousands of model evaluations on several tabular datasets. We find that the fairness-accuracy Pareto frontier achieved by postprocessing contains all other methods we were feasibly able to evaluate. In doing so, we address two common methodological errors that have confounded previous observations. One relates to the comparison of methods with different unconstrained base models. The other concerns methods achieving different levels of constraint relaxation. At the heart of our study is a simple idea we call unprocessing that roughly corresponds to the inverse of postprocessing. Unprocessing allows for a direct comparison of methods using different underlying models and levels of relaxation.

José Pombal, André F. Cruz, João Bravo et al.

In recent years, machine learning algorithms have become ubiquitous in a multitude of high-stakes decision-making applications. The unparalleled ability of machine learning algorithms to learn patterns from data also enables them to incorporate biases embedded within. A biased model can then make decisions that disproportionately harm certain groups in society -- limiting their access to financial services, for example. The awareness of this problem has given rise to the field of Fair ML, which focuses on studying, measuring, and mitigating unfairness in algorithmic prediction, with respect to a set of protected groups (e.g., race or gender). However, the underlying causes for algorithmic unfairness still remain elusive, with researchers divided between blaming either the ML algorithms or the data they are trained on. In this work, we maintain that algorithmic unfairness stems from interactions between models and biases in the data, rather than from isolated contributions of either of them. To this end, we propose a taxonomy to characterize data bias and we study a set of hypotheses regarding the fairness-accuracy trade-offs that fairness-blind ML algorithms exhibit under different data bias settings. On our real-world account-opening fraud use case, we find that each setting entails specific trade-offs, affecting fairness in expected value and variance -- the latter often going unnoticed. Moreover, we show how algorithms compare differently in terms of accuracy and fairness, depending on the biases affecting the data. Finally, we note that under specific data bias conditions, simple pre-processing interventions can successfully balance group-wise error rates, while the same techniques fail in more complex settings.

Florian E. Dorner, Yatong Chen, André F. Cruz et al.

Test-time scaling aims to improve language model performance by leveraging additional compute during inference. Many works have empirically studied techniques such as Best-of-N (BoN) and Rejection Sampling (RS) that make use of a verifier to enable test-time scaling. However, to date there is little theoretical understanding of how verifier imperfection affects performance -- a gap we address in this work. Specifically, we prove that the instance-level accuracy of these methods is precisely characterized by the geometry of the verifier's ROC curve. Our theory has two important takeaways, confirmed by experiments with Qwen and LLama models on GSM8K and MATH500. First, RS outperforms BoN for fixed compute, while both methods converge to the same accuracy in the infinite-compute limit. Second, it is generally impossible to predict the high-compute performance of either method based on observations in the low-compute regime.

André F. Cruz, Pedro Saleiro, Catarina Belém et al.

Considerable research effort has been guided towards algorithmic fairness but real-world adoption of bias reduction techniques is still scarce. Existing methods are either metric- or model-specific, require access to sensitive attributes at inference time, or carry high development or deployment costs. This work explores the unfairness that emerges when optimizing ML models solely for predictive performance, and how to mitigate it with a simple and easily deployed intervention: fairness-aware hyperparameter optimization (HO). We propose and evaluate fairness-aware variants of three popular HO algorithms: Fair Random Search, Fair TPE, and Fairband. We validate our approach on a real-world bank account opening fraud case-study, as well as on three datasets from the fairness literature. Results show that, without extra training cost, it is feasible to find models with 111% mean fairness increase and just 6% decrease in performance when compared with fairness-blind HO.

João Bento, Pedro Saleiro, André F. Cruz et al.

Although recurrent neural networks (RNNs) are state-of-the-art in numerous sequential decision-making tasks, there has been little research on explaining their predictions. In this work, we present TimeSHAP, a model-agnostic recurrent explainer that builds upon KernelSHAP and extends it to the sequential domain. TimeSHAP computes feature-, timestep-, and cell-level attributions. As sequences may be arbitrarily long, we further propose a pruning method that is shown to dramatically decrease both its computational cost and the variance of its attributions. We use TimeSHAP to explain the predictions of a real-world bank account takeover fraud detection RNN model, and draw key insights from its explanations: i) the model identifies important features and events aligned with what fraud analysts consider cues for account takeover; ii) positive predicted sequences can be pruned to only 10% of the original length, as older events have residual attribution values; iii) the most recent input event of positive predictions only contributes on average to 41% of the model's score; iv) notably high attribution to client's age, suggesting a potential discriminatory reasoning, later confirmed as higher false positive rates for older clients.

André F. Cruz, Pedro Saleiro, Catarina Belém et al.

Considerable research effort has been guided towards algorithmic fairness but there is still no major breakthrough. In practice, an exhaustive search over all possible techniques and hyperparameters is needed to find optimal fairness-accuracy trade-offs. Hence, coupled with the lack of tools for ML practitioners, real-world adoption of bias reduction methods is still scarce. To address this, we present Fairband, a bandit-based fairness-aware hyperparameter optimization (HO) algorithm. Fairband is conceptually simple, resource-efficient, easy to implement, and agnostic to both the objective metrics, model types and the hyperparameter space being explored. Moreover, by introducing fairness notions into HO, we enable seamless and efficient integration of fairness objectives into real-world ML pipelines. We compare Fairband with popular HO methods on four real-world decision-making datasets. We show that Fairband can efficiently navigate the fairness-accuracy trade-off through hyperparameter optimization. Furthermore, without extra training cost, it consistently finds configurations attaining substantially improved fairness at a comparatively small decrease in predictive accuracy.