Generative AI Pub 🎨 Image AI 👁 1 ⏱ 1 min read

28.7%: The Luna Footnote Behind OpenAI’s Decisions API

On OpenAI’s own broken-tool test, GPT-6 Luna guessed instead of flagging the failure in 28.7% of cases. The Decisions API runs on Luna.Continue reading on Generative AI »

Source: Generative AI Pub — Published — Category: Image AI

🔗 Read full article on Generative AI Pub →
🏷️ Topics: GPTOpenAI

More from Generative AI Pub today

Meta’s Muse AI and the End of User Privacy 3h ago AI Did Not Break Out of Its Sandbox. Someone Left the Firewall Open. 4h ago Muse vs Claude: Where Does Muse Actually Win? 4h ago

More in Image AI

Generative AI Pub3h ago

Meta’s Muse AI and the End of User Privacy
Generative AI Pub4h ago

AI Did Not Break Out of Its Sandbox. Someone Left the Firewall Open.
Generative AI Pub4h ago

Muse vs Claude: Where Does Muse Actually Win?
Generative AI Pub5h ago

I Benchmarked 27 LLM Configurations Across Two Machines and a Zero-Retention Cloud Tier
Generative AI Pub5h ago

How to get OpenAI Dots on Claude
Generative AI Pub5h ago

GPT-6 Sol vs Claude Opus 5.5: Same Week, Different Bet