A Comparative Analysis of LLM Memorization at Statistical and Internal Levels: Cross-Model Commonalities and Model-Specific Signatures

• URL: http://arxiv.org/abs/2603.21658v1
• Date: Mon, 23 Mar 2026 07:35:19 GMT
• Title: A Comparative Analysis of LLM Memorization at Statistical and Internal Levels: Cross-Model Commonalities and Model-Specific Signatures
• Abstract summary: Memorization is a fundamental component of intelligence for both humans and LLMs.<n>Previous studies focus on a single model series, leading to isolated observations among series.<n>In this study, we collect multiple model series and analyze their shared or unique memorization behavior.
• Score: 23.846451631333256
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Memorization is a fundamental component of intelligence for both humans and LLMs. However, while LLM performance scales rapidly, our understanding of memorization lags. Due to limited access to the pre-training data of LLMs, most previous studies focus on a single model series, leading to isolated observations among series, making it unclear which findings are general or specific. In this study, we collect multiple model series (Pythia, OpenLLaMa, StarCoder, OLMo1/2/3) and analyze their shared or unique memorization behavior at both the statistical and internal levels, connecting individual observations while showing new findings. At the statistical level, we reveal that the memorization rate scales log-linearly with model size, and memorized sequences can be further compressed. Further analysis demonstrated a shared frequency and domain distribution pattern for memorized sequences. However, different models also show individual features under the above observations. At the internal level, we find that LLMs can remove certain injected perturbations, while memorized sequences are more sensitive. By decoding middle layers and attention head ablation, we revealed the general decoding process and shared important heads for memorization. However, the distribution of those important heads differs between families, showing a unique family-level feature. Through bridging various experiments and revealing new findings, this study paves the way for a universal and fundamental understanding of memorization in LLM.

Related papers

• What Matters to an LLM? Behavioral and Computational Evidences from Summarization [9.582572639590508]
  Large Language Models (LLMs) are now state-of-the-art at summarization, yet the internal notion of importance that drives their information selections remains hidden.<n>We propose to investigate this by combining behavioral and computational analyses.
  arXiv  Detail & Related papers  (2026-01-31T02:23:30Z)
• How and Why LLMs Generalize: A Fine-Grained Analysis of LLM Reasoning from Cognitive Behaviors to Low-Level Patterns [51.02752099869218]
  Large Language Models (LLMs) display strikingly different generalization behaviors.<n>We introduce a novel benchmark that decomposes reasoning into atomic core skills.<n>We show that RL-tuned models maintain more stable behavioral profiles and resist collapse in reasoning skills, whereas SFT models exhibit sharper drift and overfit to surface patterns.
  arXiv  Detail & Related papers  (2025-12-30T08:16:20Z)
• If an LLM Were a Character, Would It Know Its Own Story? Evaluating Lifelong Learning in LLMs [55.8331366739144]
  We introduce LIFESTATE-BENCH, a benchmark designed to assess lifelong learning in large language models (LLMs)<n>Our fact checking evaluation probes models' self-awareness, episodic memory retrieval, and relationship tracking, across both parametric and non-parametric approaches.
  arXiv  Detail & Related papers  (2025-03-30T16:50:57Z)
• Skewed Memorization in Large Language Models: Quantification and Decomposition [23.097554469918357]
  Memorization in Large Language Models (LLMs) poses privacy and security risks, as models may unintentionally reproduce sensitive or copyrighted data.<n>This paper examines memorization in supervised fine-tuning (SFT), exploring its relationships with training duration, dataset size, and inter-sample similarity.
  arXiv  Detail & Related papers  (2025-02-03T09:23:53Z)
• Analyzing Memorization in Large Language Models through the Lens of Model Attribution [11.295483963637217]
  Large Language Models (LLMs) are prevalent in modern applications but often memorize training data, leading to privacy breaches and copyright issues.<n>We investigate memorization from an architectural lens by analyzing how attention modules at different layers impact its memorization and generalization.
  arXiv  Detail & Related papers  (2025-01-09T09:00:32Z)
• Generalization v.s. Memorization: Tracing Language Models' Capabilities Back to Pretraining Data [76.90128359866462]
  We introduce an extended concept of memorization, distributional memorization, which measures the correlation between the output probabilities and the pretraining data frequency.<n>This study demonstrates that memorization plays a larger role in simpler, knowledge-intensive tasks, while generalization is the key for harder, reasoning-based tasks.
  arXiv  Detail & Related papers  (2024-07-20T21:24:40Z)
• A Multi-Perspective Analysis of Memorization in Large Language Models [10.276594755936529]
  Large Language Models (LLMs) show unprecedented performance in various fields. LLMs can generate the same content used to train them. This research comprehensively discussed memorization from various perspectives.
  arXiv  Detail & Related papers  (2024-05-19T15:00:50Z)
• Dive into the Chasm: Probing the Gap between In- and Cross-Topic Generalization [66.4659448305396]
  This study analyzes various LMs with three probing-based experiments to shed light on the reasons behind the In- vs. Cross-Topic generalization gap. We demonstrate, for the first time, that generalization gaps and the robustness of the embedding space vary significantly across LMs.
  arXiv  Detail & Related papers  (2024-02-02T12:59:27Z)
• Exploring Memorization in Fine-tuned Language Models [53.52403444655213]
  We conduct the first comprehensive analysis to explore language models' memorization during fine-tuning across tasks. Our studies with open-sourced and our own fine-tuned LMs across various tasks indicate that memorization presents a strong disparity among different fine-tuning tasks. We provide an intuitive explanation of this task disparity via sparse coding theory and unveil a strong correlation between memorization and attention score distribution.
  arXiv  Detail & Related papers  (2023-10-10T15:41:26Z)
• An Empirical Study of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning [70.48605869773814]
  Catastrophic forgetting (CF) is a phenomenon that occurs in machine learning when a model forgets previously learned information.<n>This study empirically evaluates the forgetting phenomenon in large language models during continual instruction tuning.
  arXiv  Detail & Related papers  (2023-08-17T02:53:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.