Cite
Venhoff, Constantin, et al. Understanding Reasoning in Thinking Language Models via Steering Vectors. arXiv:2506.18167, arXiv, 24 June 2025. arXiv.org, https://doi.org/10.48550/arXiv.2506.18167.
Metadata
Title: Understanding Reasoning in Thinking Language Models via Steering Vectors Authors: Constantin Venhoff, Iván Arcuschin, Philip Torr, Arthur Conmy, Neel Nanda Cite key: venhoff2025Understanding
Links
Abstract
Recent advances in large language models (LLMs) have led to the development of thinking language models that generate extensive internal reasoning chains before producing responses. While these models achieve improved performance, controlling their reasoning processes remains challenging. This work presents a steering approach for thinking LLMs by analyzing and manipulating specific reasoning behaviors in DeepSeek-R1-Distill models. Through a systematic experiment on 500 tasks across 10 diverse categories, we identify several reasoning behaviors exhibited by thinking models, including expressing uncertainty, generating examples for hypothesis validation, and backtracking in reasoning chains. We demonstrate that these behaviors are mediated by linear directions in the model’s activation space and can be controlled using steering vectors. By extracting and applying these vectors, we provide a method to modulate specific aspects of the model’s reasoning process, such as its tendency to backtrack or express uncertainty. Our approach offers practical tools for steering reasoning processes in thinking models in a controlled and interpretable manner. We validate our steering method using three DeepSeek-R1-Distill models, demonstrating consistent control across different model architectures.
Notes
From Obsidian
(As notes and annotations from Zotero are one-way synced, this section include a link to another note within Obsidian to host further notes)
Understanding-Reasoning-in-Thinking-Language-Models-via-Steering-Vectors
From Zotero
(one-way sync from Zotero)
Annotations
Highlighting colour codes
Link to original
- Note: highlights for quicker reading or comments stemmed from reading the paper but might not be too related to the paper
- External Insight: Insights from other works but was mentioned in the paper
- Question/Critic: questions or comments on the content of paper
- Claim: what the paper claims to have found/achieved
- Finding: new knowledge presented by the paper
- Important: anything interesting enough (findings, insights, ideas, etc.) that’s worth remembering
From Zotero
(one-way sync from Zotero)