New benchmark and analysis reveal LLM cross-lingual summarization challenges

Researchers have introduced a new benchmark, MEA, for multi-target cross-lingual summarization (MTXLS) across 24 languages. Their analysis of large language models (LLMs) reveals that translation and summarization processes emerge jointly in later layers, rather than as separate stages. To improve MTXLS quality, they developed an inference-time method that uses hidden representations from English summarization to guide generation. AI

IMPACT Highlights limitations in LLM cross-lingual capabilities and proposes a method to improve performance.

RANK_REASON The cluster contains an academic paper detailing a new benchmark and analysis of LLM behavior. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• LLM

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →