analiza aiqbench
The introduction of Agents' Last Exam (ALE) addresses a critical evaluation gap by shifting focus from synthetic benchmarks to long-horizon, economically valuable workflows verified by real-world outcomes. With a current average pass rate of just 2.6% across mainstream configurations, ALE reveals that existing AI agents are far from saturated in professional domains, providing a necessary metric for assessing genuine GDP-relevant impact rather than leaderboard scores. This living benchmark, developed with over 250 industry experts, offers a robust framework for measuring sustained performance in non-physical sectors defined by federal occupational taxonomies.
Agents' Last Exam
arXiv:2606.05405v1 Announce Type: new Abstract: Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue that this gap is largely an evaluation problem: widely used benchmarks lack sustained performance measurement on real and economically valuable workflows. This paper introduces Agents' Last Exam (ALE), a benchmark designed to evaluate AI agents on long-horizon, economically valuable, real-world tasks with verifiable outcomes. Developed in collaboration with 250+ industry experts, ALE covers non-physical industries defined with reference to O*NET / SOC 2018 (the U.S. federal occupational taxonomy). It is organized around a task taxonomy with 55 subfields grouped into 13 industry clusters covering 1K+ tasks. Current results show that the hardest tier remains far from saturated: across mainstream harness and backbone configurations, the average full pass rate is 2.6%. ALE is designed as a living benchmark: its task pool grows continuously as new workflows and industries are onboarded. More broadly, ALE is intended not merely as another leaderboard, but as an instrument for closing the gap between benchmark success and GDP-relevant impact.
et al. (208 additional authors not shown)
Abstract:Recent AI systems have achieved strong results on a wide range of benchmarks, yet these gains have not translated into economically meaningful deployment across many professional domains. We argue that this gap is largely an evaluation problem: widely used benchmarks lack sustained performance measurement on real and economically valuable workflows. This paper introduces Agents' Last Exam (ALE), a benchmark designed to evaluate AI agents on long-horizon, economically valuable, real-world tasks with verifiable outcomes. Developed in collaboration with 250+ industry experts, ALE covers non-physical industries defined with reference to O*NET / SOC 2018 (the U.S. federal occupational taxonomy). It is organized around a task taxonomy with 55 subfields grouped into 13 industry clusters covering 1K+ tasks. Current results show that the hardest tier remains far from saturated: across mainstream harness and backbone configurations, the average full pass rate is 2.6%. ALE is designed as a living benchmark: its task pool grows continuously as new workflows and industries are onboarded. More broadly, ALE is intended not merely as another leaderboard, but as an instrument for closing the gap between benchmark success and GDP-relevant impact.
Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
arXiv:2606.05405 [cs.AI]
(or arXiv:2606.05405v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2606.05405 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 214 cuvinte