Publications

Black-Box Access is Insufficient for Rigorous Al Audits
Stephen Casper
Carson Ezell
Charlotte Siegmann
Noam Kolt
Taylor Lynn Curtis
Benjamin Bucknall
Andreas Haupt
Kevin Wei
Jérémy Scheurer
Marius Hobbhahn
Lee Sharkey
Satyapriya Krishna
Marvin Von Hagen
Silas Alberti
Alan Chan
Qinyi Sun
Michael Gerovitch
David Bau
Max Tegmark
David Krueger … (see 1 more)
Dylan Hadfield-Menell
External audits of AI systems are increasingly recognized as a key mechanism for AI governance. The effectiveness of an audit, however, depe… (see more)nds on the degree of access granted to auditors. Recent audits of state-of-the-art AI systems have primarily relied on black-box access, in which auditors can only query the system and observe its outputs. However, white-box access to the system's inner workings (e.g., weights, activations, gradients) allows an auditor to perform stronger attacks, more thoroughly interpret models, and conduct fine-tuning. Meanwhile, outside-the-box access to training and deployment information (e.g., methodology, code, documentation, data, deployment details, findings from internal evaluations) allows auditors to scrutinize the development process and design more targeted evaluations. In this paper, we examine the limitations of black-box audits and the advantages of white- and outside-the-box audits. We also discuss technical, physical, and legal safeguards for performing these audits with minimal security risks. Given that different forms of access can lead to very different levels of evaluation, we conclude that (1) transparency regarding the access and methods used by auditors is necessary to properly interpret audit results, and (2) white- and outside-the-box access allow for substantially more scrutiny than black-box access alone.
2024-06-04
The 2024 ACM Conference on Fairness, Accountability, and Transparency (published)
doi.org
arxiv.org
Characterizing and Classifying Developer Forum Posts with their Intentions
Xingfang Wu
Eric Laufer
Heng Li
Foutse Khomh
Santhosh Srinivasan
Jayden Luo
With the rapid growth of the developer community, the amount of posts on online technical forums has been growing rapidly, which poses diffi… (see more)culties for users to filter useful posts and find important information. Tags provide a concise feature dimension for users to locate their interested posts and for search engines to index the most relevant posts according to the queries. However, most tags are only focused on the technical perspective (e.g., program language, platform, tool). In most cases, forum posts in online developer communities reveal the author's intentions to solve a problem, ask for advice, share information, etc. The modeling of the intentions of posts can provide an extra dimension to the current tag taxonomy. By referencing previous studies and learning from industrial perspectives, we create a refined taxonomy for the intentions of technical forum posts. Through manual labeling and analysis on a sampled post dataset extracted from online forums, we understand the relevance between the constitution of posts (code, error messages) and their intentions. Furthermore, inspired by our manual study, we design a pre-trained transformer-based model to automatically predict post intentions. The best variant of our intention prediction framework, which achieves a Micro F1-score of 0.589, Top 1-3 accuracy of 62.6% to 87.8%, and an average AUC of 0.787, outperforms the state-of-the-art baseline approach. Our characterization and automated classification of forum posts regarding their intentions may help forum maintainers or third-party tool developers improve the organization and retrieval of posts on technical forums. We have released our annotated dataset and codes in our supplementary material package.
2024-06-04
Empirical Software Engineering (published)
doi.org
arxiv.org
Individual Brain Charting dataset extension, third release for movie watching and retinotopy data
Ana Lúısa Pinho
Hugo Richard
Ana Fernanda Ponce
Michael Eickenberg
Alexis Amadon
Elvis Dopgima Dohmatob
Isabelle Denghien
Juan Jesús Torre
Swetha Shankar
Himanshu Aggarwal
Alexis Thual
Thomas Chapalain
Chantal Ginisty
Séverine Becuwe-Desmidt
Séverine Roger
Yann Lecomte
Valérie Berland
Laurence Laurier
Véronique Joly-Testault
Gaëlle Médiouni-Cloarec … (see 6 more)
Christine Doublé
Bernadette Martins
Gael Varoquaux
Stanislas Dehaene
Lucie Hertz-Pannier
Bertrand Thirion
2024-06-04
Scientific Data (published)
doi.org
IrokoBench: A New Benchmark for African Languages in the Age of Large Language Models
David Ifeoluwa Adelani
Jessica Ojo
Israel Abebe Azime
Zhuang Yun Jian
Jesujoba Oluwadara Alabi
Xuanli He
Millicent Ochieng
Sara Hooker
Andiswa Bukula
En-Shiun Annie Lee
Chiamaka Ijeoma Chukwuneke
Happy Buzaaba
Blessing Kudzaishe Sibanda
Godson Kalipe
Jonathan Mukiibi
Salomon Kabongo
Foutse Yuehgoh
M. Setaka
Lolwethu Ndolela
Nkiruka Bridget Odu … (see 6 more)
Rooweither Mabuya
Shamsuddeen Hassan Muhammad
Salomey Osei
Sokhar Samb
Tadesse Kebede Guge
Pontus Stenetorp
Despite the widespread adoption of Large language models (LLMs), their remarkable capabilities remain limited to a few high-resource languag… (see more)es. Additionally, many low-resource languages (e.g. African languages) are often evaluated only on basic text classification tasks due to the lack of appropriate or comprehensive benchmarks outside of high-resource languages. In this paper, we introduce IrokoBench -- a human-translated benchmark dataset for 16 typologically-diverse low-resource African languages covering three tasks: natural language inference~(AfriXNLI), mathematical reasoning~(AfriMGSM), and multi-choice knowledge-based QA~(AfriMMLU). We use IrokoBench to evaluate zero-shot, few-shot, and translate-test settings~(where test sets are translated into English) across 10 open and four proprietary LLMs. Our evaluation reveals a significant performance gap between high-resource languages~(such as English and French) and low-resource African languages. We observe a significant performance gap between open and proprietary models, with the highest performing open model, Aya-101 only at 58\% of the best-performing proprietary model GPT-4o performance. Machine translating the test set to English before evaluation helped to close the gap for larger models that are English-centric, like LLaMa 3 70B. These findings suggest that more efforts are needed to develop and adapt LLMs for African languages.
2024-06-04
ArXiv (preprint)
doi.org
arxiv.org
Machine Learning Data Practices through a Data Curation Lens: An Evaluation Framework
Eshta Bhardwaj
Harshit Gujral
Siyi Wu
Ciara Zogheib
Tegan Maharaj
Christoph Becker
Studies of dataset development in machine learning call for greater attention to the data practices that make model development possible and… (see more) shape its outcomes. Many argue that the adoption of theory and practices from archives and data curation fields can support greater fairness, accountability, transparency, and more ethical machine learning. In response, this paper examines data practices in machine learning dataset development through the lens of data curation. We evaluate data practices in machine learning as data curation practices. To do so, we develop a framework for evaluating machine learning datasets using data curation concepts and principles through a rubric. Through a mixed-methods analysis of evaluation results for 25 ML datasets, we study the feasibility of data curation principles to be adopted for machine learning data work in practice and explore how data curation is currently performed. We find that researchers in machine learning, which often emphasizes model development, struggle to apply standard data curation principles. Our findings illustrate difficulties at the intersection of these fields, such as evaluating dimensions that have shared terms in both fields but non-shared meanings, a high degree of interpretative flexibility in adapting concepts without prescriptive restrictions, obstacles in limiting the depth of data curation expertise needed to apply the rubric, and challenges in scoping the extent of documentation dataset creators are responsible for. We propose ways to address these challenges and develop an overall framework for evaluation that outlines how data curation concepts and methods can inform machine learning data practices.
2024-06-04
The 2024 ACM Conference on Fairness, Accountability, and Transparency (published)
doi.org
arxiv.org
Meta’s AI translation model embraces overlooked languages
David I. Adelani
2024-06-04
Nature (unknown)
doi.org
Noisy Data Visualization using Functional Data Analysis
Haozhe Chen
Andres Felipe Duque Correa
Guy Wolf
Kevin R. Moon
Data visualization via dimensionality reduction is an important tool in exploratory data analysis. However, when the data are noisy, many ex… (see more)isting methods fail to capture the underlying structure of the data. The method called Empirical Intrinsic Geometry (EIG) was previously proposed for performing dimensionality reduction on high dimensional dynamical processes while theoretically eliminating all noise. However, implementing EIG in practice requires the construction of high-dimensional histograms, which suffer from the curse of dimensionality. Here we propose a new data visualization method called Functional Information Geometry (FIG) for dynamical processes that adapts the EIG framework while using approaches from functional data analysis to mitigate the curse of dimensionality. We experimentally demonstrate that the resulting method outperforms a variant of EIG designed for visualization in terms of capturing the true structure, hyperparameter robustness, and computational speed. We then use our method to visualize EEG brain measurements of sleep activity.
2024-06-04
ArXiv (preprint)
doi.org
arxiv.org
A Robot Walks into a Bar: Can Language Models Serve as Creativity SupportTools for Comedy? An Evaluation of LLMs' Humour Alignment with Comedians
Piotr Mirowski
J Christopher Love
Juliette Love
Kory Mathewson
Shakir Mohamed
2024-06-04
The 2024 ACM Conference on Fairness, Accountability, and Transparency (published)
doi.org
arxiv.org
Towards Geographic Inclusion in the Evaluation of Text-to-Image Models
Melissa Hall
Samuel J. Bell
Candace Ross
Adina Williams
Michal Drozdzal
Adriana Romero
Rapid progress in text-to-image generative models coupled with their deployment for visual content creation has magnified the importance of … (see more)thoroughly evaluating their performance and identifying potential biases. In pursuit of models that generate images that are realistic, diverse, visually appealing, and consistent with the given prompt, researchers and practitioners often turn to automated metrics to facilitate scalable and cost-effective performance profiling. However, commonly-used metrics often fail to account for the full diversity of human preference; often even in-depth human evaluations face challenges with subjectivity, especially as interpretations of evaluation criteria vary across regions and cultures. In this work, we conduct a large, cross-cultural study to study how much annotators in Africa, Europe, and Southeast Asia vary in their perception of geographic representation, visual appeal, and consistency in real and generated images from state-of-the art public APIs. We collect over 65,000 image annotations and 20 survey responses. We contrast human annotations with common automated metrics, finding that human preferences vary notably across geographic location and that current metrics do not fully account for this diversity. For example, annotators in different locations often disagree on whether exaggerated, stereotypical depictions of a region are considered geographically representative. In addition, the utility of automatic evaluations is dependent on assumptions about their set-up, such as the alignment of feature extractors with human perception of object similarity or the definition of"appeal"captured in reference datasets used to ground evaluations. We recommend steps for improved automatic and human evaluations.
2024-06-04
The 2024 ACM Conference on Fairness, Accountability, and Transparency (published)
doi.org
arxiv.org
Visibility into AI Agents
Alan Chan
Carson Ezell
Max Kaufmann
Kevin Wei
Lewis Hammond
Herbie Bradley
Emma Bluemke
Nitarshan Rajkumar
David Krueger
Noam Kolt
Lennart Heim
Markus Anderljung
Increased delegation of commercial, scientific, governmental, and personal activities to AI agents -- systems capable of pursuing complex go… (see more)als with limited supervision -- may exacerbate existing societal risks and introduce new risks. Understanding and mitigating these risks involves critically evaluating existing governance structures, revising and adapting these structures where needed, and ensuring accountability of key stakeholders. Information about where, why, how, and by whom certain AI agents are used, which we refer to as visibility, is critical to these objectives. In this paper, we assess three categories of measures to increase visibility into AI agents: agent identifiers, real-time monitoring, and activity logging. For each, we outline potential implementations that vary in intrusiveness and informativeness. We analyze how the measures apply across a spectrum of centralized through decentralized deployment contexts, accounting for various actors in the supply chain including hardware and software service providers. Finally, we discuss the implications of our measures for privacy and concentration of power. Further work into understanding the measures and mitigating their negative impacts can help to build a foundation for the governance of AI agents.
2024-06-04
The 2024 ACM Conference on Fairness, Accountability, and Transparency (published)
doi.org
arxiv.org
Milnor-Myerson Games and The Principles of Artificial Principal-Agent Problems
Manfred Diaz
Joel Z Leibo
Liam Paull
In this paper, we introduce Milnor-Myerson games, a multiplayer interaction structure at the core of machine learning (ML), to shed light on… (see more) the fundamental principles and implications the artificial principal-agent problem has had in landmark ML results like AlphaGo and large language models (LLMs).
2024-06-03
rl-conference.cc/RLC/2024/Workshop/Finding_the_Frame (poster)
openreview.net
openreview.net
From Feature Visualization to Visual Circuits: Effect of Adversarial Model Manipulation
Geraldin Nanfack
Michael Eickenberg
Eugene Belilovsky
Understanding the inner working functionality of large-scale deep neural networks is challenging yet crucial in several high-stakes applicat… (see more)ions. Mechanistic inter- pretability is an emergent field that tackles this challenge, often by identifying human-understandable subgraphs in deep neural networks known as circuits. In vision-pretrained models, these subgraphs are usually interpreted by visualizing their node features through a popular technique called feature visualization. Recent works have analyzed the stability of different feature visualization types under the adversarial model manipulation framework. This paper starts by addressing limitations in existing works by proposing a novel attack called ProxPulse that simultaneously manipulates the two types of feature visualizations. Surprisingly, when analyzing these attacks under the umbrella of visual circuits, we find that visual circuits show some robustness to ProxPulse. We, therefore, introduce a new attack based on ProxPulse that unveils the manipulability of visual circuits, shedding light on their lack of robustness. The effectiveness of these attacks is validated using pre-trained AlexNet and ResNet-50 models on ImageNet.
2024-06-02
ArXiv (preprint)
doi.org
arxiv.org