ainewsblitz.com

Breaking

OpenAI GPT-5.6 Sol, Terra, and Luna post 88-93% hallucination rates on AA-Omniscience

  • Foundation Models
  • Research & Papers

OpenAI's GPT-5.6 family is posting hallucination rates of 88 to 93 percent on the AA-Omniscience knowledge benchmark, even as flagship GPT-5.6 Sol remains among the most accurate models on the same 6,000-question test.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,575 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year