BURMESE-SAN: Burmese NLP Benchmark for Evaluating Large Language Models

arXiv:2602.18788v2 Announce Type: replace Abstract: We introduce BURMESE-SAN, the first holistic benchmark that systematically evaluates large language models (LLMs) for Burmese across three core NLP competencies: understanding (NLU), reasoning (NLR), and generation (NLG). BUR...

🔗 Read more: https://arxiv.org/abs/2602.18788

#News #Policy #AI #Biology #Academic