Facts: First amplify correlations and then slice to discover bias

S Yenamandra, P Ramesh… - Proceedings of the …, 2023 - openaccess.thecvf.com
Computer vision datasets frequently contain spurious correlations between task-relevant
labels and (easy to learn) latent task-irrelevant attributes (eg context). Models trained on …

Recite, reconstruct, recollect: Memorization in LMs as a multifaceted phenomenon

US Prashanth, A Deng, K O'Brien, J SV… - ar** out the Memorisation-Generalisation Continuum in Neural Machine Translation
V Dankers, I Titov, D Hupkes - ar**s from
your dataset but never learn some others. Yet, memorisation is not easily expressed as a …

Resource-lean transfer methods for cross-lingual information retrieval

RM Litschko - 2024 - madoc.bib.uni-mannheim.de
Cross-Lingual Information Retrieval (CLIR) is the task of finding relevant documents written
in a language different from the query language. Neural machine translation systems and …