Recent research in mechanistic interpretability has attempted to
reverse...
Many NLP datasets have been found to contain shortcuts: simple decision ...
Diversity is an important criterion for many areas of machine learning (...
Many datasets have been created for training reading comprehension model...
Petroni et al. (2019) demonstrated that it is possible to retrieve world...
Scientific article summarization is challenging: large, annotated corpor...
Scientific article summarization is challenging: large, annotated corpor...
We introduce a new syntax-aware model for dependency-based semantic role...
Building robust online content recommendation systems requires learning
...