The500Feed.Live
Everything going on in AI - updated daily from 500+ sources
📄 ResearchJune 1, 2026
CARTE: A Benchmark for Mapping Language Model Knowledge Across France
We introduce CARTE 1 (Culturally Anchored Regional-Territorial Evaluation), a multiplechoice benchmark for evaluating the ability of large language models (LLMs) to perform fine-grained reasoning over geographically grounded and regionally differentiated knowledge within France. While prior benchmar...
Read Original Article →