Across 28 models, knowledge breadth tracks intelligence at Pearson r=0.86
teortaxesTex · x · 2026-10-29
Reconstructing a scatter chart of 28 models, the author finds a strong correlation (Pearson r=0.86, r²=0.742, OLS y≈1.196x−25.98) between the AA-Omniscience Index and Intelligence Index, concluding there's "no such thing as a small under-post-trained frontier model."
Positive residual outliers (more knowledge than trend predicts) include Muse Spark Open (Δ+22.3), Grok 4.7 xhigh (Δ+16.2), Claude Fable 5.1 (Δ+13.7), and Gemini 3.8 Flash (Δ+12.8); negative outliers include MiMo-V2.6-Pro (Δ−14.2), DeepSeek V4.1 Flash 1017 (Δ−11.9), and Mistral Large 4 Preview (Δ−11.7).
More from Models
- Rime launches prism speech-to-speech model: 300ms response, 4x faster than OpenAI — EXM7777 · 2026-10-29
- Claim: Chinese labs now distilling from OpenAI's Astra after Anthropic blockade — bindureddy · 2026-10-29
- "Desired oververbosity 5": Claude spots OpenAI distillation fingerprints in DeepSeek reasoning traces — zeeshanp_ · 2026-10-29
- Qwen4 leak points to 1M-token context window, Qwen4-27B expected tomorrow — brainExploded99 · 2026-10-29
- Muse-Glimmer flops in hands-on test: user gives up at 40K tokens, switches to DeepSeek — techmago · 2026-10-29
- Pedro Domingos mocks NYT: the brain runs one dimension per synapse — a trillion — pmddomingos · 2026-10-29