Analysis Finds Thinking Machines' Model Swears When Benchmarks Stump It

Vals AI analyzed 2,655 benchmark runs of Thinking Machines' Inkling 2 Small and found the model swore in about 50 of them, typically when faced with unsolvable problems.

2026-11-07 ~ 2026-11-07 · 2 related posts

• Inkling 2 Small swears in 50 of 2,655 benchmark runs when debugging fails — LiTianleli · 2026-11-07
• Thinking Machines' Inkling 2 Small swears while debugging — 50 of 2,655 benchmark runs — SonglinYang4 · 2026-11-07