Fooling LIME and SHAP

Dylan Slack, Sophie Hilgard, Emily Jia, +2

Proceedings of the AAAI/ACM Conference on AI Ethics and Society

Artificial IntelligenceEngineeringMachine LearningAi SafetyInformation Forensics +15

Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods

Dylan Slack, Sophie Hilgard, Emily Jia, +2

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningInformation SecurityVerification +23

Explaining machine learning models with interactive natural language conversations using TalkToModel

Dylan Slack, Satyapriya Krishna, Himabindu Lakkaraju, +1

Nature Machine Intelligence

Artificial IntelligenceEngineeringMachine LearningSpoken Dialog SystemNatural Language Processing +15

Assessing the Local Interpretability of Machine Learning Models

Dylan Slack, Sorelle A. Friedler, Carlos Scheidegger, +1

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningLocal InterpretabilityAi Safety +15

Fairness warnings and fair-MAML

Dylan Slack, Sorelle A. Friedler, Emile Givental

Artificial IntelligenceFairness WarningsMachine LearningData ScienceEngineering +11

How can we fool LIME and SHAP? Adversarial Attacks on Post hoc Explanation Methods.

Dylan Slack, Sophie Hilgard, Emily Jia, +2

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningInformation SecurityVerification +18

Rethinking Explainability as a Dialogue: A Practitioner's Perspective

Himabindu Lakkaraju, Dylan Slack, Yuxin Chen, +2

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningCommunicationNatural Language Processing +17

Active meta-learning for predicting and selecting perovskite crystallization experiments

Venkateswaran Shekar, Gareth Nicholas, Mansoor Ani Najeeb, +10

The Journal of Chemical Physics

Artificial IntelligenceAutonomous Experimentation SystemsEngineeringMachine LearningMeta-learning +18

Reliable Post hoc Explanations: Modeling Uncertainty in Explainability

Dylan Slack, Sophie Hilgard, Sameer Singh, +1

arXiv (Cornell University)

Artificial IntelligenceEngineeringMachine LearningBayesian VersionsCommunication +19

Post Hoc Explanations of Language Models Can Improve Language Models

Satyapriya Krishna, Jiaqi Ma, Dylan Slack, +3