3 papers
cs.LG2024
EXAGREE: Mitigating Explanation Disagreement with Stakeholder-Aligned Models
Sichao Li, Tommy Liu, Quanling Deng +1
Conflicting explanations, arising from different attribution methods or model internals, limit the adoption of machine learning models in safety-critical domains. We turn this disa…
cs.LG2024
Practical Attribution Guidance for Rashomon Sets
Sichao Li, Amanda S. Barnard, Quanling Deng
Different prediction models might perform equally well (Rashomon set) in the same task, but offer conflicting interpretations and conclusions about the data. The Rashomon effect in…
cs.LG2024
Diverse Explanations From Data-Driven and Domain-Driven Perspectives in the Physical Sciences
Sichao Li, Xin Wang, Amanda Barnard
Machine learning methods have been remarkably successful in material science, providing novel scientific insights, guiding future laboratory experiments, and accelerating materials…