coastalcph/fair-rationales
Explainability methods are used to benchmark the extent to which model predictions align with human rationales i.e., are 'right for the right reasons'. Previous work has failed to acknowledge, however, that what counts as a rationale is sometimes subjective. This paper presents what we think is a first of its kind, a collection of human rationale annotations augmented with the annotators demographic information.
Upload fair-rationales.py
Upload fair-rationales.py
Update README.md
Upload fair-rationales.py
Upload 3 files
Upload fair-rationales.py
Upload fair-rationales.py
Upload fair-rationales.py
Delete fair_rationales.py
Upload 4 files
Upload fair_rationales.py
Delete fair-rationales.zip
Upload 4 files
Upload 2 files
Upload fair-rationales.zip
Upload fair_rationales.py
Upload fair_rationales.py
Upload 2 files
Upload fair_rationales.py
Upload fair-rationales.zip
Upload fair-rationales.zip
Delete fair-rationales.zip
Upload fair_rationales.py
Upload fair_rationales.py
Upload 2 files
Delete fair-rationales.zip
Upload fair-rationales.zip
Delete fair-rationales.zip
Delete SST_annot.json
Delete cose_annot.json
Delete dynasent_annot.json
Upload 2 files
Upload fair_rationales.py
Upload 4 files
Update README.md
Update README.md
initial commit