no code implementations • 3 May 2024 • Minh Duc Bui, Katharina von der Wense
Current natural language processing (NLP) research tends to focus on only one or, less frequently, two dimensions - e. g., performance, privacy, fairness, or efficiency - at a time, which may lead to suboptimal conclusions and often overlooking the broader goal of achieving trustworthy NLP.
no code implementations • 30 Apr 2024 • Minh Duc Bui, Fabian David Schmidt, Goran Glavaš, Katharina von der Wense
We further find that KD yields larger gains over pretraining from scratch when the data must be repeated under the fixed computation budget.
no code implementations • 21 Mar 2024 • Enora Rice, Ali Marashian, Luke Gessler, Alexis Palmer, Katharina von der Wense
Canonical morphological segmentation is the process of analyzing words into the standard (aka underlying) forms of their constituent morphemes.
no code implementations • 31 Jan 2024 • Sagi Shaier, Lawrence E Hunter, Katharina von der Wense
Prior work has uncovered a set of common problems in state-of-the-art context-based question answering (QA) systems: a lack of attention to the context when the latter conflicts with a model's parametric knowledge, little robustness to noise, and a lack of consistency with their answers.
no code implementations • 31 Jan 2024 • Sagi Shaier, Kevin Bennett, Lawrence E Hunter, Katharina von der Wense
(RQ2) Do models' absolute scores differ between the two approaches?
no code implementations • 16 Oct 2023 • Sagi Shaier, Lawrence E. Hunter, Katharina von der Wense
In this opinion piece, we argue that LMs in their current state will never be fully trustworthy in critical settings and suggest a possible novel strategy to handle this issue: by building LMs such that can cite their sources - i. e., point a user to the parts of their training data that back up their outputs.
1 code implementation • 16 Oct 2023 • Sagi Shaier, Kevin Bennett, Lawrence Hunter, Katharina von der Wense
State-of-the-art question answering (QA) models exhibit a variety of social biases (e. g., with respect to sex or race), generally explained by similar issues in their training data.