# Towards Understanding Distilled Reasoning Models: A Representational Approach

arXiv: https://arxiv.org/abs/2503.03730 · HTML: https://mlsanity.com/paper/2503.03730/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning.
- Builds on 29 of its 38 references that are in the map.
- First posted on arXiv 5 Mar 2025.

## Placed in

- LLM reasoning / [Prompting and chain-of-thought reasoning](https://mlsanity.com/lr/prompting-and-chain-of-thought-reasoning/)
- Knowledge distillation / [Knowledge distillation for LLMs](https://mlsanity.com/distill/knowledge-distillation-for-llms/)

## Builds on (most cited first, 29 of 29)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [Sparse Autoencoders Find Highly Interpretable Features in Language Models](https://mlsanity.com/paper/2309.08600/) (2023)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [Steering Language Models With Activation Engineering](https://mlsanity.com/paper/2308.10248/) (2023)
- [The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets](https://mlsanity.com/paper/2310.06824/) (2023)
- [Scaling and evaluating sparse autoencoders](https://mlsanity.com/paper/2406.04093/) (2024)
- [Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2](https://mlsanity.com/paper/2408.05147/) (2024)
- [Language Models Represent Space and Time](https://mlsanity.com/paper/2310.02207/) (2023)
- [Function Vectors in Large Language Models](https://mlsanity.com/paper/2310.15213/) (2023)
- [Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders](https://mlsanity.com/paper/2407.14435/) (2024)
- [Model compression via distillation and quantization](https://mlsanity.com/paper/1802.05668/) (2018)
- [A Survey on Knowledge Distillation of Large Language Models](https://mlsanity.com/paper/2402.13116/) (2024)
- [Not All Language Model Features Are One-Dimensionally Linear](https://mlsanity.com/paper/2405.14860/) (2024)
- [Implicit Representations of Meaning in Neural Language Models](https://mlsanity.com/paper/2106.00737/) (2021)
- [Long Short-Term Memory-Networks for Machine Reading](https://mlsanity.com/paper/1601.06733/) (2016)
- [The Geometry of Categorical and Hierarchical Concepts in Large Language Models](https://mlsanity.com/paper/2406.01506/) (2024)
- [Omnigrok: Grokking Beyond Algorithmic Data](https://mlsanity.com/paper/2210.01117/) (2022)
- [Language Models Use Trigonometry to Do Addition](https://mlsanity.com/paper/2502.00873/) (2025)
- [Improving Activation Steering in Language Models with Mean-Centring](https://mlsanity.com/paper/2312.03813/) (2023)
- [Extending Activation Steering to Broad Skills and Multiple Behaviours](https://mlsanity.com/paper/2403.05767/) (2024)
- [Transcoders Beat Sparse Autoencoders for Interpretability](https://mlsanity.com/paper/2501.18823/) (2025)
- [Memory Injections: Correcting Multi-Hop Reasoning Failures during Inference in Transformer-Based Language Models](https://mlsanity.com/paper/2309.05605/) (2023)
- [Opening the AI black box: program synthesis via mechanistic interpretability](https://mlsanity.com/paper/2402.05110/) (2024)
- [Investigating Representation Universality: Case Study on Genealogical Representations](https://mlsanity.com/paper/2410.08255/) (2024)
- [Harmonic Loss Trains Interpretable AI Models](https://mlsanity.com/paper/2502.01628/) (2025)

## Built on by (most cited first, 26 of 26)

- [Speculative Thinking: Enhancing Small-Model Reasoning with Large Model Guidance at Inference Time](https://mlsanity.com/paper/2504.12329/) (2025)
- [Mitigating Overthinking in Large Reasoning Models via Manifold Steering](https://mlsanity.com/paper/2505.22411/) (2025)
- [Base Models Know How to Reason, Thinking Models Learn When](https://mlsanity.com/paper/2510.07364/) (2025)
- [Emergent Coordination in Multi-Agent Language Models](https://mlsanity.com/paper/2510.05174/) (2025)
- [Trade-offs in Large Reasoning Models: An Empirical Analysis of Deliberative and Adaptive Reasoning over Foundational Capabilities](https://mlsanity.com/paper/2503.17979/) (2025)
- [On Reasoning Strength Planning in Large Reasoning Models](https://mlsanity.com/paper/2506.08390/) (2025)
- [Reasoning or Retrieval? A Study of Answer Attribution on Large Reasoning Models](https://mlsanity.com/paper/2509.24156/) (2025)
- [Lexical Hints of Accuracy in LLM Reasoning Chains](https://mlsanity.com/paper/2508.15842/) (2025)
- [Demystifying Hybrid Thinking: Can LLMs Truly Switch Between Think and No-Think?](https://mlsanity.com/paper/2510.12680/) (2025)
- [Large Reasoning Models Are (Not Yet) Multilingual Latent Reasoners](https://mlsanity.com/paper/2601.02996/) (2026)
- [Crosscoding Through Time: Tracking Emergence & Consolidation Of Linguistic Representations Throughout LLM Pretraining](https://mlsanity.com/paper/2509.05291/) (2025)
- [Internal states before wait modulate reasoning patterns](https://mlsanity.com/paper/2510.04128/) (2025)
- [Towards a Mechanistic Understanding of Large Reasoning Models: A Survey of Training, Inference, and Failures](https://mlsanity.com/paper/2601.19928/) (2026)
- [Distilling the Essence: Efficient Reasoning Distillation via Sequence Truncation](https://mlsanity.com/paper/2512.21002/) (2025)
- [CodeReasoner: Enhancing the Code Reasoning Ability with Reinforcement Learning](https://mlsanity.com/paper/2507.17548/) (2025)
- [Rank-1 LoRAs Encode Interpretable Reasoning Signals](https://mlsanity.com/paper/2511.06739/) (2025)
- [Latent Concept Disentanglement in Transformer-based Language Models](https://mlsanity.com/paper/2506.16975/) (2025)
- [Why Does Reinforcement Learning Generalize? A Feature-Level Mechanistic Study of Post-Training in Large Language Models](https://mlsanity.com/paper/2604.25011/) (2026)
- [Distractor Injection Attacks on Large Reasoning Models: Characterization and Defense](https://mlsanity.com/paper/2510.16259/) (2025)
- [Transcoder Adapters for Reasoning-Model Diffing](https://mlsanity.com/paper/2602.20904/) (2026)
- [Fine-Tuning Small Reasoning Models for Quantum Field Theory](https://mlsanity.com/paper/2604.18936/) (2026)
- [Understanding On-Policy Distillation: A Mechanistic Interpretability Perspective via Sparse Crosscoders](https://mlsanity.com/paper/2609.35210/) (2026)
- [Distilled Circuits: A Mechanistic Study of Internal Restructuring in Knowledge Distillation](https://mlsanity.com/paper/2505.10822/) (2025)
- [Enhancing Reasoning Capabilities in SLMs with Reward Guided Dataset Distillation](https://mlsanity.com/paper/2507.00054/) (2025)
- [fmxcoders: Factorized Masked Crosscoders for Cross-Layer Feature Discovery](https://mlsanity.com/paper/2605.09438/) (2026)
- [A Survey on the Linear Representation Hypothesis](https://mlsanity.com/paper/2609.22695/) (2026)

## Abstract (arXiv)

In this paper, we investigate how model distillation impacts the development of reasoning features in large language models (LLMs). To explore this, we train a crosscoder on Qwen-series models and their fine-tuned variants. Our results suggest that the crosscoder learns features corresponding to various types of reasoning, including self-reflection and computation verification. Moreover, we observe that distilled models contain unique reasoning feature directions, which could be used to steer the model into over-thinking or incisive-thinking mode. In particular, we perform analysis on four specific reasoning categories: (a) self-reflection, (b) deductive reasoning, (c) alternative reasoning, and (d) contrastive reasoning. Finally, we examine the changes in feature geometry resulting from the distillation process and find indications that larger distilled models may develop more structured representations, which correlate with enhanced distillation performance. By providing insights into how distillation modifies the model, our study contributes to enhancing the transparency and reliability of AI systems.