OpenAI briefly halved GPT-6 Astra's hallucination rate in post-launch benchmark edit
The company also boosted its own math scores and cut Anthropic's, raising questions about 'benchmaxxing' in AI evaluation.
By Turki Al-Mutairi·· 3 min

Loading the Newsroom
Curating from trusted global sources…
4 briefings · “evA”
The company also boosted its own math scores and cut Anthropic's, raising questions about 'benchmaxxing' in AI evaluation.
Paul Griffiths insists Dubai's global hub ambitions are intact after the Iran war cut passenger traffic by nearly a third. Here is why he is doubling down.
The fast-fashion giant's shares fell on its first day of trading, a stark signal that the trade advantages and buzz that fueled its rise have evaporated.
A Harvard valedictorian used Claude to hit a 1939 breakthrough, but the missing “why” is the real problem.