Ahead of AI (Sebastian Raschka)· Sebastian Raschka, PhD·AI score58
Raschka explains how LLMs control reasoning effort settings
Controlling Reasoning Effort in LLMs
Summary
Sebastian Raschka's newsletter article explains how reasoning-effort settings in models like OpenAI's GPT-5.6 are trained and implemented. It describes effort-conditioned supervised fine-tuning and reinforcement learning with token-length penalties, and compares disclosed recipes from six open-weight models including DeepSeek V4, Kimi K2.5, GLM-5, Qwen3 and Nemotron 3 Ultra. Higher effort generally means longer responses and higher cost.
This site provides summaries and links to the originals only.
Source: Ahead of AI (Sebastian Raschka) · magazine.sebastianraschka.com