arXiv subjects

Tongshuang Wu

Publications and source records attributed to Tongshuang Wu.

75 records · Page 5Linked to original sources

arXiv2021

Does the Whole Exceed its Parts? The Effect of AI Explanations on Complementary Team Performance

Many researchers motivate explainable AI with studies showing that human-AI team performance on decision-making tasks improves when the AI explains its recommendations. However, prior studies observed improvements from explanations only when the AI, alone, outperformed both the human and the best team. Can explanations help lead to complementary performance, where team accuracy is higher than either the human or the AI working solo? We conduct mixed-method user studies on three datasets, where an AI with accuracy comparable to humans helps participants solve a task (explaining itself in some conditions). While we observed complementary improvements from AI augmentation, they were not increased by explanations. Rather, explanations increased the chance that humans will accept the AI's recommendation, regardless of its correctness. Our result poses new challenges for human-centered AI: Can we develop explanatory approaches that encourage appropriate trust in AI, and therefore help generate (or improve) complementary performance?

cs.AI↗

arXiv2020

Beyond Accuracy: Behavioral Testing of NLP models with CheckList

Although measuring held-out accuracy has been the primary approach to evaluate generalization, it often overestimates the performance of NLP models, while alternative approaches for evaluating models either focus on individual tasks or on specific behaviors. Inspired by principles of behavioral testing in software engineering, we introduce CheckList, a task-agnostic methodology for testing NLP models. CheckList includes a matrix of general linguistic capabilities and test types that facilitate comprehensive test ideation, as well as a software tool to generate a large and diverse number of test cases quickly. We illustrate the utility of CheckList with tests for three tasks, identifying critical failures in both commercial and state-of-art models. In a user study, a team responsible for a commercial sentiment analysis model found new and actionable bugs in an extensively tested model. In another user study, NLP practitioners with CheckList created twice as many tests, and found almost three times as many bugs as users without it.

cs.CL↗

arXiv2019

Technology-Enabled Disinformation: Summary, Lessons, and Recommendations

Technology is increasingly used -- unintentionally (misinformation) or intentionally (disinformation) -- to spread false information at scale, with potentially broad-reaching societal effects. For example, technology enables increasingly realistic false images and videos, and hyper-personal targeting means different people may see different versions of reality. This report is the culmination of a PhD-level special topics course (https://courses.cs.washington.edu/courses/cse599b/18au/) in Computer Science & Engineering at the University of Washington's Paul G. Allen School in the fall of 2018. The goals of this course were to study (1) how technologies and today's technical platforms enable and support the creation and spread of such mis- and disinformation, as well as (2) how technical approaches could be used to mitigate these issues. In this report, we summarize the space of technology-enabled mis- and disinformation based on our investigations, and then surface our lessons and recommendations for technologists, researchers, platform designers, policymakers, and users.

cs.CY↗