# Reasoning Models Can be Accurately Pruned Via Chain-of-Thought Reconstruction

arXiv: https://arxiv.org/abs/2509.12464 · HTML: https://mlsanity.com/paper/2509.12464/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning.
- Builds on 10 of its 21 references that are in the map.
- First posted on arXiv 15 Sep 2025.

## Placed in

- LLM reasoning / [Test-time compute and efficient thinking](https://mlsanity.com/lr/test-time-compute-and-efficient-thinking/)

## Builds on (most cited first, 10 of 10)

- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Tree of Thoughts: Deliberate Problem Solving with Large Language Models](https://mlsanity.com/paper/2305.10601/) (2023)
- [On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes](https://mlsanity.com/paper/2306.13649/) (2023)
- [A Simple and Effective Pruning Approach for Large Language Models](https://mlsanity.com/paper/2306.11695/) (2023)
- [Accelerating Sparse Deep Neural Networks](https://mlsanity.com/paper/2104.08378/) (2021)
- [CodeElo: Benchmarking Competition-level Code Generation of LLMs with Human-comparable Elo Ratings](https://mlsanity.com/paper/2501.01257/) (2025)
- [Is C4 Dataset Optimal for Pruning? An Investigation of Calibration Data for LLM Pruning](https://mlsanity.com/paper/2410.07461/) (2024)
- [ALPS: Improved Optimization for Highly Sparse One-Shot Pruning for Large Language Models](https://mlsanity.com/paper/2406.07831/) (2024)
- [When Reasoning Meets Compression: Understanding the Effects of LLMs Compression on Large Reasoning Models](https://mlsanity.com/paper/2504.02010/) (2025)
- [PPC-GPT: Federated Task-Specific Compression of Large Language Models via Pruning and Chain-of-Thought Distillation](https://mlsanity.com/paper/2502.15857/) (2025)

## Built on by (most cited first, 4 of 4)

- [OBC-Prune: Outcome-Based Calibration for Large Reasoning Model Pruning](https://mlsanity.com/paper/2609.17890/) (2026)
- [Hardware-Native Joint Sparse-Quantization for Trillion-Scale Mixture-of-Experts](https://mlsanity.com/paper/2610.02241/) (2026)
- [Dynamics as Code: On Model Compression via Dynamic System](https://mlsanity.com/paper/2610.11115/) (2026)
- [Unlocking Structured Pruning for Reasoning Language Models via Attention-Preserving MLP Trimming](https://mlsanity.com/paper/2610.19121/) (2026)

## Abstract (arXiv)

Reasoning language models such as DeepSeek-R1 produce long chain-of-thought traces during inference time which make them costly to deploy at scale. We show that using compression techniques such as neural network pruning produces greater performance loss than in typical language modeling tasks, and in some cases can make the model slower since they cause the model to produce more thinking tokens but with worse performance. We show that this is partly due to the fact that standard LLM pruning methods often focus on input reconstruction, whereas reasoning is a decode-dominated task. We introduce a simple, drop-in fix: during pruning we jointly reconstruct activations from the input and the model's on-policy chain-of-thought traces. This "Reasoning-Aware Compression" (RAC) integrates seamlessly into existing pruning workflows such as SparseGPT, and boosts their performance significantly. Code reproducing the results in the paper can be found at: https://github.com/RyanLucas3/RAC