English

CLINB: A Climate Intelligence Benchmark for Foundational Models

Artificial Intelligence 2025-11-18 v1 Computation and Language

Abstract

Evaluating how Large Language Models (LLMs) handle complex, specialized knowledge remains a critical challenge. We address this through the lens of climate change by introducing CLINB, a benchmark that assesses models on open-ended, grounded, multimodal question answering tasks with clear requirements for knowledge quality and evidential support. CLINB relies on a dataset of real users' questions and evaluation rubrics curated by leading climate scientists. We implement and validate a model-based evaluation process and evaluate several frontier models. Our findings reveal a critical dichotomy. Frontier models demonstrate remarkable knowledge synthesis capabilities, often exhibiting PhD-level understanding and presentation quality. They outperform "hybrid" answers curated by domain experts assisted by weaker models. However, this performance is countered by failures in grounding. The quality of evidence varies, with substantial hallucination rates for references and images. We argue that bridging this gap between knowledge synthesis and verifiable attribution is essential for the deployment of AI in scientific workflows and that reliable, interpretable benchmarks like CLINB are needed to progress towards building trustworthy AI systems.

Keywords

Cite

@article{arxiv.2511.11597,
  title  = {CLINB: A Climate Intelligence Benchmark for Foundational Models},
  author = {Michelle Chen Huebscher and Katharine Mach and Aleksandar Stanić and Markus Leippold and Ben Gaiarin and Zeke Hausfather and Elisa Rawat and Erich Fischer and Massimiliano Ciaramita and Joeri Rogelj and Christian Buck and Lierni Sestorain Saralegui and Reto Knutti},
  journal= {arXiv preprint arXiv:2511.11597},
  year   = {2025}
}

Comments

Questions, system prompt and model judge prompts available here: https://www.kaggle.com/datasets/deepmind/clinb-questions

R2 v1 2026-07-01T07:37:57.475Z