Activation Control for Efficiently Eliciting Long Chain-of-thought Ability of Language Models

Understand

Despite the remarkable reasoning performance, eliciting the long chain-of-thought (CoT) ability in large language models (LLMs) typically requires costly reinforcement learning or supervised fine-tuning on high-quality distilled data.

• We investigate the internal mechanisms behind this capability and show that a small set of high-impact activations in the last few layers largely governs long-form reasoning attributes, such as output length and self-reflection.
• By simply amplifying these activations and inserting "wait" tokens, we can invoke the long CoT ability without any training, resulting in significantly increased self-reflection rates and accuracy.
• Moreover, we find that the activation dynamics follow predictable trajectories, with a sharp rise after special tokens and a subsequent exponential decay.