JasonChen91/Earth-Iron
(ICLR'26) EarthSE: A Benchmark for Evaluating Earth Scientific Exploration Capability of LLMs Updates/News π π© News (2026-01-26) EarthSE has been accepted by ICLR 2026 π. Abstract Advancements in Large Language Models (LLMs) drive interest in scientific applications, necessitating specialized benchmarks such as Earth science. Existing benchmarks either present a general science focus devoid of Earth science specificity orβ¦ See the full description on the dataset page: https://huggingface.co/datasets/JasonChen91/Earth-Iron.
021
Duplicate from ai-earth/Earth-Iron
