Report: 'GPT-6 Luna (max)' scores 18.3 at n=3, degradation allegedly confirmed — unverified
PawelHuryn · x · 2026-09-24
Pawel Huryn claims "confirmed: GPT-6 Luna (max): 18.3 for n=3. The degradation was real," adding that at n=1 it's only slightly better than a locally runnable Qwen3.8-27B. Unverified and unofficial — if real, it points to a significant degradation issue under repeated sampling/eval settings.
Related event: Leaks Map GPT-6 Codenames to GPT-5.6 Versions(3 posts)→
More from Models
- antirez backs low-bit DeepSeek 4.1 Flash over Qwen 27B dense on Apple M5 Macs — antirez · 2026-11-12
- Gemini re-renders images as text when vision is removed, dropping only 6% WSR — zacharynado · 2026-11-12
- Google's Physical-Engineering's Last Exam: all 14 frontier models score below 50% — dl_weekly · 2026-11-12
- Commit leak reveals unreleased Gemini 4 variants: Argon, Hydrogen and Helix (unconfirmed) — BlackHC · 2026-11-11
- Rewiring just 89 of 442,368 neurons in Qwen3-8B matches full SFT for a persona — maxsloef · 2026-11-11
- Unsloth ships Qwen3.8-27B Dynamic V3.1 GGUFs with 5-10% accuracy gains — danielhanchen · 2026-11-11