Humanity's Last Exam

TL;DR

Humanity's Last Exam (HLE) is a comprehensive, multi-modal benchmark designed to challenge large language models with questions that are difficult for current models but accessible to humans, aiming to better measure true AI capabilities.

Contribution

The paper introduces HLE, a globally developed, expert-curated benchmark with 2,500 questions across diverse subjects, filling the gap left by existing benchmarks that models now easily surpass.

Findings

LLMs achieve low accuracy on HLE, exposing limitations.

HLE questions cannot be quickly answered via internet retrieval.

HLE provides a more accurate measure of LLM capabilities.

Abstract

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 2,500 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai.