# The Reasoning-Memorization Interplay in Language Models Is Mediated by a Single Direction

arXiv: https://arxiv.org/abs/2503.23084 · HTML: https://mlsanity.com/paper/2503.23084/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 37 of its 67 references that are in the map.
- First posted on arXiv 29 Mar 2025.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 30 of 37)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Mistral 7B](https://mlsanity.com/paper/2310.06825/) (2023)
- [Code Llama: Open Foundation Models for Code](https://mlsanity.com/paper/2308.12950/) (2023)
- [Gemma 2: Improving Open Language Models at a Practical Size](https://mlsanity.com/paper/2408.00118/) (2024)
- [Refusal in Language Models Is Mediated by a Single Direction](https://mlsanity.com/paper/2406.11717/) (2024)
- [Toy Models of Superposition](https://mlsanity.com/paper/2209.10652/) (2022)
- [The Linear Representation Hypothesis and the Geometry of Large Language Models](https://mlsanity.com/paper/2311.03658/) (2023)
- [The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets](https://mlsanity.com/paper/2310.06824/) (2023)
- [Measuring Faithfulness in Chain-of-Thought Reasoning](https://mlsanity.com/paper/2307.13702/) (2023)
- [Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets](https://mlsanity.com/paper/2201.02177/) (2022)
- [Language Models are Multilingual Chain-of-Thought Reasoners](https://mlsanity.com/paper/2210.03057/) (2022)
- [2 OLMo 2 Furious](https://mlsanity.com/paper/2501.00656/) (2024)
- [Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks](https://mlsanity.com/paper/2307.02477/) (2023)
- [Proving Test Set Contamination in Black Box Language Models](https://mlsanity.com/paper/2310.17623/) (2023)
- [Embers of Autoregression: Understanding Large Language Models Through the Problem They are Trained to Solve](https://mlsanity.com/paper/2309.13638/) (2023)
- [Do Large Language Models Latently Perform Multi-Hop Reasoning?](https://mlsanity.com/paper/2402.16837/) (2024)
- [Linear Representations of Sentiment in Large Language Models](https://mlsanity.com/paper/2310.15154/) (2023)
- [Causal Parrots: Large Language Models May Talk Causality But Are Not Causal](https://mlsanity.com/paper/2308.13067/) (2023)
- [Emergent and Predictable Memorization in Large Language Models](https://mlsanity.com/paper/2304.11158/) (2023)
- [On Memorization of Large Language Models in Logical Reasoning](https://mlsanity.com/paper/2410.23123/) (2024)
- [Physics of Language Models: Part 2.1, Grade-School Math and the Hidden Reasoning Process](https://mlsanity.com/paper/2407.20311/) (2024)
- [Grokked Transformers are Implicit Reasoners: A Mechanistic Journey to the Edge of Generalization](https://mlsanity.com/paper/2405.15071/) (2024)
- [Hopping Too Late: Exploring the Limitations of Large Language Models on Multi-Hop Queries](https://mlsanity.com/paper/2406.12775/) (2024)
- [Attention Heads of Large Language Models: A Survey](https://mlsanity.com/paper/2409.03752/) (2024)
- [Robust LLM safeguarding via refusal feature adversarial training](https://mlsanity.com/paper/2409.20089/) (2024)

## Built on by (most cited first, 21 of 21)

- [AlphaSteer: Learning Refusal Steering with Principled Null-Space Constraint](https://mlsanity.com/paper/2506.07022/) (2025)
- [Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models](https://mlsanity.com/paper/2601.14004/) (2026)
- [MetaFaith: Faithful Natural Language Uncertainty Expression in LLMs](https://mlsanity.com/paper/2505.24858/) (2025)
- [Spurious Rewards Paradox: Mechanistically Understanding How RLVR Activates Memorization Shortcuts in LLMs](https://mlsanity.com/paper/2601.11061/) (2026)
- [On Reasoning Strength Planning in Large Reasoning Models](https://mlsanity.com/paper/2506.08390/) (2025)
- [Detecting (Un)answerability in Large Language Models with Linear Directions](https://mlsanity.com/paper/2509.22449/) (2025)
- [The Landscape of Memorization in LLMs: Mechanisms, Measurement, and Mitigation](https://mlsanity.com/paper/2507.05578/) (2025)
- [MLLMEraser: Achieving Test-Time Unlearning in Multimodal Large Language Models through Activation Steering](https://mlsanity.com/paper/2510.04217/) (2025)
- [Hypothesis-Driven Feature Manifold Analysis in LLMs via Supervised Multi-Dimensional Scaling](https://mlsanity.com/paper/2510.01025/) (2025)
- [Stochastic Chameleons: Irrelevant Context Hallucinations Reveal Class-Based (Mis)Generalization in LLMs](https://mlsanity.com/paper/2505.22630/) (2025)
- [Turning the Spell Around: Lightweight Alignment Amplification via Rank-One Safety Injection](https://mlsanity.com/paper/2508.20766/) (2025)
- [Diagnosing Memorization in Chain-of-Thought Reasoning, One Token at a Time](https://mlsanity.com/paper/2508.02037/) (2025)
- [ThoughtProbe: Classifier-Guided LLM Thought Space Exploration via Probing Representations](https://mlsanity.com/paper/2510.27355/) (2025)
- [Eliciting Chain-of-Thought in Base LLMs via Gradient-Based Representation Optimization](https://mlsanity.com/paper/2511.19131/) (2025)
- [Adaptive Multi-Agent Reasoning via Automated Workflow Generation](https://mlsanity.com/paper/2507.14393/) (2025)
- [Memory Dial: A Training Framework for Controllable Memorization in Language Models](https://mlsanity.com/paper/2604.05074/) (2026)
- [The Master Key Hypothesis: Unlocking Cross-Model Capability Transfer via Linear Subspace Alignment](https://mlsanity.com/paper/2604.06377/) (2026)
- [Multilingual Idioms in Sentences and Conversations Across High-, Medium-, and Low-Resource Languages](https://mlsanity.com/paper/2606.02147/) (2026)
- [IsoSci: A Benchmark of Isomorphic Cross-Domain Science Problems for Evaluating Reasoning versus Knowledge Retrieval in LLMs](https://mlsanity.com/paper/2607.01431/) (2026)
- [Beyond Fixed Directions: Adaptive Representation Analysis of Reasoning and Memorization in LLMs](https://mlsanity.com/paper/2608.21919/) (2026)
- [How Do Linear Probes Emerge? A Circuit-Tracing Framework with Concept-Targeted Attribution](https://mlsanity.com/paper/2608.27510/) (2026)

## Abstract (arXiv)

Large language models (LLMs) excel on a variety of reasoning benchmarks, but previous studies suggest they sometimes struggle to generalize to unseen questions, potentially due to over-reliance on memorized training examples. However, the precise conditions under which LLMs switch between reasoning and memorization during text generation remain unclear. In this work, we provide a mechanistic understanding of LLMs' reasoning-memorization dynamics by identifying a set of linear features in the model's residual stream that govern the balance between genuine reasoning and memory recall. These features not only distinguish reasoning tasks from memory-intensive ones but can also be manipulated to causally influence model performance on reasoning tasks. Additionally, we show that intervening in these reasoning features helps the model more accurately activate the most relevant problem-solving capabilities during answer generation. Our findings offer new insights into the underlying mechanisms of reasoning and memory in LLMs and pave the way for the development of more robust and interpretable generative AI systems.