Evaluating Large Language Models for Multilingual Abstractive Summarization: A Comprehensive Study of English and Bengali
arXiv:2610.19003v1 Announce Type: new Abstract: Abstractive summarization remains challenging across languages, where linguistic quality and factual consistency may diverge, particularly in low-resource settings. We present a standardized zero-shot benchmark of eight closed-source and open-weight L…
Read the full story at arXiv cs.CL ↗