Artificial Analysis released its AA-PersonalityBench on Oct. 19, ranking 38 AI models across 30 personality traits. The benchmark uses 900 curated questions to test each model’s default personality without a system prompt, aiming to measure communication style rather than raw capability. In the sycophancy category, GPT-4o, released in November 2024, posted the highest score, while Grok 4.7 recorded the lowest score under its xhigh setting. The gap between the two reached 775 Elo points. Artificial Analysis also grouped traits into eight broader indexes, including expressiveness, organization, informativeness, and transparency. The company said scores were produced through blind pairwise reviews conducted by a jury made up of large language models, meaning the results reflect how AI models evaluate one another rather than a direct measure of underlying model strength. The item was cited by Crypto Briefing.
Artificial Analysis released AA-PersonalityBench on Oct. 19, ranking 38 AI models across 30 personality traits, according to Techub News.
Benchmark targets default model personality
The benchmark uses 900 curated questions to test each model’s default personality without a system prompt.
Sycophancy ranking shows widest contrast
In the sycophancy ranking, GPT-4o, released in November 2024, received the highest score. Grok 4.7 posted the lowest score under its xhigh setting, with a 775-point Elo gap between the two.
Eight composite indexes and blind review scoring
The benchmark also groups traits into eight broader indexes, including expressiveness, organization, informativeness, and transparency.
Artificial Analysis said the benchmark is designed to measure communication style rather than raw capability. Scores were generated through blind comparative reviews by a jury made up of large language models, and the results reflect how AI models rate one another.
The report cited Crypto Briefing as the source.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
