MapEval, a 700-question benchmark spanning textual, API, and visual map tasks, finds no foundation model exceeds 67% accuracy while humans score at least 82%.
We can see that all models struggled, with the best accuracy being only 51.06%
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
MapEval: A Map-Based Evaluation of Geo-Spatial Reasoning in Foundation Models
MapEval, a 700-question benchmark spanning textual, API, and visual map tasks, finds no foundation model exceeds 67% accuracy while humans score at least 82%.