Grok takes instructions too literally, listing icons instead of using them
Daniel_Farinax · x · 2026-08-17
Users report that Grok has been overly literal lately. When told to "use the same icons as in the mobile app," it listed the icon names under a heading instead of applying them, a behavior repeated in various scenarios.
More from Models
- Reviewer claims Gemini 3.7 Flash outperforms predecessor — cgarciae88 · 2026-08-17
- Qwen3.8-27B-Abliterated-MLX Model Released on Hugging Face — PocketAiHub · 2026-08-17
- Review: Qwen 2.5 72B excels at coding but fails logic benchmarks with severe hallucinations — Afinetheorem · 2026-08-17
- Research finds Claude text watermarking alters meaning — petrusenko_max · 2026-08-17
- Community discussion: Minimax H3 training suffers due to distilled model nature — Beneficial_Toe_2347 · 2026-08-17
- OpenAI training models to write superhumanly secure code, says Brockman — tekbog · 2026-08-17