Papers by Muhammad Arham Mirza
PakBBQ: A Culturally Adapted Bias Benchmark for QA (2025.emnlp-main)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are widely adopted in language processing applications, but they often perpetuate harmful societal biases. |
| Approach: | They propose a culturally and regionally adapted extension of the original Bias Benchmark for Question Answering dataset to address this gap. |
| Outcome: | The proposed model gains 12% accuracy with disambiguation and stronger counter bias behaviors in Urdu than in English. |