Test: Style-Adjusting Skills Affect Model Chain of Thought
eliebakouch · x · 2026-08-24
An experiment reveals that using a skill to adjust a model's output style also impacts its Chain of Thought. Instructing the model not to apply the style during reasoning did not alter the outcome. Additionally, the author notes that DeepSeek V4 Pro almost never uses em dashes.
Related event: Style Skills Quietly Rewrite Model Chain-of-Thought(2 posts)→
More from Models
- New inference engine 'e' hits 10,500 tok/s on a single DGX Station — brandon_galang · 2026-10-16
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Opus 5 might be a very deep model or use looping, commentator speculates — scaling01 · 2026-10-16
- GPT-6.1 Sol audit finds 6% of arXiv papers had false main theorems, down to 1.5% recently — nitarshan · 2026-10-16
- Nvidia-backed post-training pushes Nemotron 3.5 VL past frontier on dermatology diagnosis — LoganGrasby · 2026-10-16
- Gemini 4 Argon scores 97.5% on ARC-AGI-3, second lab above 95% — patience_cave · 2026-10-16