Papers by Günter Neumann
An Interactive Web-Interface for Visualizing the Inner Workings of the Question Answering LSTM (D18-2)
Copied to clipboard
| Challenge: | Existing visualisation methods for deep learning models are limited by their low interpretability and lack a tool for interpreting them. |
| Approach: | They propose a visualisation tool which plots heatmaps of neurons’ firings and allows a user to check the dependency between neurons and manual features. |
| Outcome: | The proposed visualisation tool plots heatmaps of neurons’ firings and allows a user to check the dependency between neurons and manual features. |
Only for the Unseen Languages, Say the Llamas: On the Efficacy of Language Adapters for Cross-lingual Transfer in English-centric LLMs (2025.acl-srw)
Copied to clipboard
| Challenge: | Most state-of-the-art large language models (LLMs) are trained mainly on English data, limiting their effectiveness on non-English, especially low-resource, languages. |
| Approach: | They train language adapters for 13 languages and evaluate their effectiveness on downstream tasks using either task adapters or in-context learning. |
| Outcome: | The proposed language adapters improve performance for languages not seen during pretraining, but provide negligible benefit for seen languages. |