Optimising Human-AI Collaboration by Learning Convincing Explanations
Not on the map. Why?
Abstract (arXiv)
Machine learning models are being increasingly deployed to take, or assist in taking, complicated and high-impact decisions, from quasi-autonomous vehicles to clinical decision support systems. This poses challenges, particularly when models have hard-to-detect failure modes and are able to take actions without oversight. In order to handle this challenge, we propose a method for a collaborative system that remains safe by having a human ultimately making decisions, while giving the model the best opportunity to convince and debate them with interpretable explanations. However, the most helpful explanation varies among individuals and may be inconsistent across stated preferences. To this end we develop an algorithm, Ardent, to efficiently learn a ranking through interaction and best assist humans complete a task. By utilising a collaborative approach, we can ensure safety and improve performance while addressing transparency and accountability concerns. Ardent enables efficient and effective decision-making by adapting to individual preferences for explanations, which we validate through extensive simulations alongside a user study involving a challenging image classification task, demonstrating consistent improvement over competing systems.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 21 of its 42 references that are in the map. First posted on arXiv 13 Nov 2023.
Builds on
21 of 42 references.
References · 21
- 2017Axiomatic Attribution for Deep Networkscited by 3,384
- 2017Learning Important Features Through Propagating Activation Differencescited by 1,458
- 2018Explaining Explanations: An Overview of Interpretability of Machine Learningcited by 508
- 2017A Tutorial on Thompson Samplingcited by 433
- 2018AI safety via debatecited by 400
- 2020Consistent Estimators for Learning to Defer to an Expertcited by 271
- 2015Cascading Bandits: Learning to Rank in the Cascade Modelcited by 77
- 2021Scalable Bayesian Inverse Reinforcement Learningcited by 56
- 2017MNL-Bandit: A Dynamic Learning Approach to Assortment Selectioncited by 50
- 2021Explaining Latent Representations with a Corpus of Examplescited by 30
- 2022Estimating and Penalizing Induced Preference Shifts in Recommender Systemscited by 29
- 2022Uncalibrated Models Can Improve Human-AI Collaborationcited by 26
9 more
- 2018Interpreting Neural Networks With Nearest Neighborscited by 18
- 2022POETREE: Interpretable Policy Learning with Adaptive Decision Treescited by 14
- 2021Inverse Contextual Bandits: Learning How Behavior Evolves over Timecited by 12
- 2018What is Interpretable? Using Machine Learning to Design Interpretable Decision-Support Systemscited by 11
- 2021The Medkit-Learn(ing) Environment: Medical Decision Modelling through Simulationcited by 11
- 2023Inverse Decision Modeling: Learning Interpretable Representations of Behaviorcited by 10
- 2020Sequential Explanations with Mental Model-Based Policiescited by 8
- 2022Inverse Online Learning: Understanding Non-Stationary and Reactionary Policiescited by 6
- 2022Practical Approaches for Fair Learning with Multitype and Multivariate Sensitive Attributescited by 3
Built on by
Citing papers per year: 2024: 1.
1 paper.