Whose Preferences? Differences in Fairness Preferences and Their Impact on the Fairness of AI Utilizing Human Feedback

There is a growing body of work on learning from human feedback to align various aspects of machine learning systems with human values and preferences. We consider the setting of fairness in content moderation, in which human feedback is used to determine how two comments -- referencing different sensitive attribute groups -- should be treated in comparison to one another. With a novel dataset collected from Prolific and MTurk, we find significant gaps in fairness preferences depending on the race, age, political stance, educational level, and LGBTQ+ identity of annotators. We also demonstrate that demographics mentioned in text have a strong influence on how users perceive individual fairness in moderation. Further, we find that differences also exist in downstream classifiers trained to predict human preferences. Finally, we observe that an ensemble, giving equal weight to classifiers trained on annotations from different demographics, performs better for different demographic intersections; compared to a single classifier that gives equal weight to each annotation.

SocialGenPod: Privacy-Friendly Generative AI Social Web Applications with Decentralised Personal Data Stores

We present SocialGenPod, a decentralised and privacy-friendly way of deploying generative AI Web applications. Unlike centralised Web and data architectures that keep user data tied to application and service providers, we show how one can use Solid -- a decentralised Web specification -- to decouple user data from generative AI applications. We demonstrate SocialGenPod using a prototype that allows users to converse with different Large Language Models, optionally leveraging Retrieval Augmented Generation to generate answers grounded in private documents stored in any Solid Pod that the user is allowed to access, directly or indirectly. SocialGenPod makes use of Solid access control mechanisms to give users full control of determining who has access to data stored in their Pods. SocialGenPod keeps all user data (chat history, app configuration, personal documents, etc) securely in the user's personal Pod; separate from specific model or application providers. Besides better privacy controls, this approach also enables portability across different services and applications. Finally, we discuss challenges, posed by the large compute requirements of state-of-the-art models, that future research in this area should address. Our prototype is open-source and available at: https://github.com/Vidminas/socialgenpod/.

FairTargetSim: An Interactive Simulator for Understanding and Explaining the Fairness Effects of Target Variable Definition

Machine learning requires defining one's target variable for predictions or decisions, a process that can have profound implications on fairness: biases are often encoded in target variable definition itself, before any data collection or training. We present an interactive simulator, FairTargetSim (FTS), that illustrates how target variable definition impacts fairness. FTS is a valuable tool for algorithm developers, researchers, and non-technical stakeholders. FTS uses a case study of algorithmic hiring, using real-world data and user-defined target variables. FTS is open-source and available at: http://tinyurl.com/ftsinterface. The video accompanying this paper is here: http://tinyurl.com/ijcaifts.

On The Truthfulness of 'Surprisingly Likely' Responses of Large Language Models

The surprisingly likely criterion in the seminal work of Prelec (the Bayesian Truth Serum) guarantees truthfulness in a game-theoretic multi-agent setting, by rewarding rational agents to maximise the expected information gain with their answers w.r.t. their probabilistic beliefs. We investigate the relevance of a similar criterion for responses of LLMs. We hypothesize that if the surprisingly likely criterion works in LLMs, under certain conditions, the responses that maximize the reward under this criterion should be more accurate than the responses that only maximize the posterior probability. Using benchmarks including the TruthfulQA benchmark and using openly available LLMs: GPT-2 and LLaMA-2, we show that the method indeed improves the accuracy significantly (for example, upto 24 percentage points aggregate improvement on TruthfulQA and upto 70 percentage points improvement on individual categories of questions).

Decentralised, Scalable and Privacy-Preserving Synthetic Data Generation

Synthetic data is emerging as a promising way to harness the value of data, while reducing privacy risks. The potential of synthetic data is not limited to privacy-friendly data release, but also includes complementing real data in use-cases such as training machine learning algorithms that are more fair and robust to distribution shifts etc. There is a lot of interest in algorithmic advances in synthetic data generation for providing better privacy and statistical guarantees and for its better utilisation in machine learning pipelines. However, for responsible and trustworthy synthetic data generation, it is not sufficient to focus only on these algorithmic aspects and instead, a holistic view of the synthetic data generation pipeline must be considered. We build a novel system that allows the contributors of real data to autonomously participate in differentially private synthetic data generation without relying on a trusted centre. Our modular, general and scalable solution is based on three building blocks namely: Solid (Social Linked Data), MPC (Secure Multi-Party Computation) and Trusted Execution Environments (TEEs). Solid is a specification that lets people store their data securely in decentralised data stores called Pods and control access to their data. MPC refers to the set of cryptographic methods for different parties to jointly compute a function over their inputs while keeping those inputs private. TEEs such as Intel SGX rely on hardware based features for confidentiality and integrity of code and data. We show how these three technologies can be effectively used to address various challenges in responsible and trustworthy synthetic data generation by ensuring: 1) contributor autonomy, 2) decentralisation, 3) privacy and 4) scalability. We support our claims with rigorous empirical results on simulated and real datasets and different synthetic data generation algorithms.

WCLD: Curated Large Dataset of Criminal Cases from Wisconsin Circuit Courts

Machine learning based decision-support tools in criminal justice systems are subjects of intense discussions and academic research. There are important open questions about the utility and fairness of such tools. Academic researchers often rely on a few small datasets that are not sufficient to empirically study various real-world aspects of these questions. In this paper, we contribute WCLD, a curated large dataset of 1.5 million criminal cases from circuit courts in the U.S. state of Wisconsin. We used reliable public data from 1970 to 2020 to curate attributes like prior criminal counts and recidivism outcomes. The dataset contains large number of samples from five racial groups, in addition to information like sex and age (at judgment and first offense). Other attributes in this dataset include neighborhood characteristics obtained from census data, detailed types of offense, charge severity, case decisions, sentence lengths, year of filing etc. We also provide pseudo-identifiers for judge, county and zipcode. The dataset will not only enable researchers to more rigorously study algorithmic fairness in the context of criminal justice, but also relate algorithmic challenges with various systemic issues. We also discuss in detail the process of constructing the dataset and provide a datasheet. The WCLD dataset is available at \url{https://clezdata.github.io/wcld/}.

Human-Guided Fair Classification for Natural Language Processing

Text classifiers have promising applications in high-stake tasks such as resume screening and content moderation. These classifiers must be fair and avoid discriminatory decisions by being invariant to perturbations of sensitive attributes such as gender or ethnicity. However, there is a gap between human intuition about these perturbations and the formal similarity specifications capturing them. While existing research has started to address this gap, current methods are based on hardcoded word replacements, resulting in specifications with limited expressivity or ones that fail to fully align with human intuition (e.g., in cases of asymmetric counterfactuals). This work proposes novel methods for bridging this gap by discovering expressive and intuitive individual fairness specifications. We show how to leverage unsupervised style transfer and GPT-3's zero-shot capabilities to automatically generate expressive candidate pairs of semantically similar sentences that differ along sensitive attributes. We then validate the generated pairs via an extensive crowdsourcing study, which confirms that a lot of these pairs align with human intuition about fairness in the context of toxicity classification. Finally, we show how limited amounts of human feedback can be leveraged to learn a similarity specification that can be used to train downstream fairness-aware models.

The Importance of Modeling Data Missingness in Algorithmic Fairness: A Causal Perspective

Training datasets for machine learning often have some form of missingness. For example, to learn a model for deciding whom to give a loan, the available training data includes individuals who were given a loan in the past, but not those who were not. This missingness, if ignored, nullifies any fairness guarantee of the training procedure when the model is deployed. Using causal graphs, we characterize the missingness mechanisms in different real-world scenarios. We show conditions under which various distributions, used in popular fairness algorithms, can or can not be recovered from the training data. Our theoretical results imply that many of these algorithms can not guarantee fairness in practice. Modeling missingness also helps to identify correct design principles for fair algorithms. For example, in multi-stage settings where decisions are made in multiple screening rounds, we use our framework to derive the minimal distributions required to design a fair algorithm. Our proposed algorithm decentralizes the decision-making process and still achieves similar performance to the optimal algorithm that requires centralization and non-recoverable distributions.

Infochain: A Decentralized System for Truthful Information Elicitation

Incentive mechanisms play a pivotal role in collecting correct and reliable information from self-interested agents. Peer-prediction mechanisms are game-theoretic mechanisms that incentivize agents for reporting the information truthfully, even when the information is unverifiable in nature. Traditionally, a trusted third party implements these mechanisms. We built Infochain, a decentralized system for information elicitation. Infochain ensures transparent, trustless and cost-efficient collection of information from self-interested agents without compromising the game-theoretical guarantees of the peer-prediction mechanisms. In this paper, we address various non-trivial challenges in implementing these mechanisms in Ethereum and provide experimental analysis.

Crowdsourcing with Fairness, Diversity and Budget Constraints

Recent studies have shown that the labels collected from crowdworkers can be discriminatory with respect to sensitive attributes such as gender and race. This raises questions about the suitability of using crowdsourced data for further use, such as for training machine learning algorithms. In this work, we address the problem of fair and diverse data collection from a crowd under budget constraints. We propose a novel algorithm which maximizes the expected accuracy of the collected data, while ensuring that the errors satisfy desired notions of fairness. We provide guarantees on the performance of our algorithm and show that the algorithm performs well in practice through experiments on real dataset.