Cultural biases in multilingual datasets pose significant challenges for their effectiveness as
global benchmarks. These biases stem not only from language but also from the cultural …

The performance differential of large language models (LLM) between languages hinders
their effective deployment in many regions, inhibiting the potential economic and societal …

English-centric large language models (LLMs) often show strong multilingual capabilities.
However, the multilingual performance of these models remains unclear and is not …

Large language models (LLMs) for African languages perform worse compared to their
performance in high-resource languages. To address this issue, we introduce AfriInstruct …

Multilingual natural language processing is getting increased attention, with numerous
models, benchmarks, and methods being released for many languages. English is often …

The current best practice to measure the performance of base Large Language Models is to
establish a multi-task benchmark that covers a range of capabilities of interest. Currently …

Evaluations of Large Language Models (LLMs) on knowledge-intensive tasks and factual
accuracy often focus on high-resource languages primarily because datasets for low …

Artificial conversational agents are deployed to assist humans in a variety of tasks. Some of
these tasks require the capability to communicate numbers as part of their internal and …

Current LLM structured pruning methods involve two steps:(1) compressing with calibration
data and (2) continued pretraining on billions of tokens to recover the lost performance. This …

Recent advancements in language models, particularly for high-resource languages, have
not been paralleled in low-resource languages spoken across Africa. This thesis addresses …