MiniMax M3.1 Flash accused of benchmaxxing after -22 point drop on updated AA Omniscience Index
rickasaurus · x · 2026-10-29
An X user suspects MiniMax M3.1 Flash's sudden standout score on Artificial Analysis's Omniscience bench is benchmaxxing: the closed benchmark only exposes 600 of 6k questions, and M3.1-Flash appeared on AA's page just 1 hour after release (vs. 1-2 days for DeepSeek), suggesting early access to the eval.
The accompanying AA-Omniscience Index v1→v1.1 chart supports the suspicion: after v1.1 removed saturated questions, MiniMax M3.1-Flash dropped -22.0 — the largest decline among 28 compared models (Qwen3.8 27B: -3.6) — consistent with overfitting to known items. Unverified allegation.
More from Models
- Odyssey-3 world models now simulate general environments in real time with higher physical fidelity — Scobleizer · 2026-10-29
- Qualia Beats Flora and GPT-6 Astra Across Three Robot-Data Benchmarks — lukas_m_ziegler · 2026-10-29
- Anthropic: removing 'think carefully' prompts cut Opus 5.5 latency from 8.2s to 2.1s — FinanceYF5 · 2026-10-29
- Anthropic releases Claude Opus 5.5 prompting guide with 11 best practices — FinanceYF5 · 2026-10-29
- Particle-1 inference goes live on OpenRouter, claiming 28-52% fewer output tokens — ycombinator · 2026-10-29
- Qwen3.8-27B EXL3 quant hits lowest KLD of 0.0002 in just 23 GB — suprjami · 2026-10-29