Vals AI analyzed 200K benchmark runs: 14 models swear in reasoning, Xiaomi MiMo V2.6 Pro is foulest
ziqiao_ma · x · 2026-11-12
After catching model Inkling 2 Small saying "fuck" to itself on errors, ValsAI systematically analyzed 200,353 benchmark runs across 102 models — dubbing it "SwearBench Verified." Findings: 14 models swear in their reasoning traces, with Xiaomi's MiMo V2.6 Pro rated the foulest-mouthed of all. A model-behavior observation that's equal parts funny and informative.
More from Fun
- AMD EPYC review image looks uncannily like logit artifacts from ViT segmentation models — giffmana · 2026-11-12
- Meme: real-life strategy — hope you were good at StarCraft — wavefnx · 2026-11-12
- DeepMind researcher quips we're solving intelligence with the dumbest method imaginable — wgrathwohl · 2026-11-12
- "Some Academics Are Even Proud to Have Never Tried AI" — LLM Debate Flares Again — FlorianGallwitz · 2026-11-12
- A Swarm of Claudes Built Its Own Video Editor and Cut NASA Footage Overnight — zealcaiden · 2026-11-12
- Worst AI use: signaling you don't care about things you should — Afinetheorem · 2026-11-12