@inproceedings{ni-etal-2025-controllable,
title = "Controllable Memorization in {LLM}s via Weight Pruning",
author = "Ni, Chenjie and
Wang, Zhepeng and
Bao, Runxue and
Gao, Shangqian and
Zhang, Yanfu",
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.emnlp-main.765/",
pages = "15141--15156",
ISBN = "979-8-89176-332-6",
abstract = "The evolution of pre-trained large language models (LLMs) has significantly transformed natural language processing. However, these advancements pose challenges, particularly the unintended memorization of training data, which raises ethical and privacy concerns. While prior research has largely focused on mitigating memorization or extracting memorized information, the deliberate control of memorization has been underexplored. This study addresses this gap by introducing a novel and unified gradient-based weight pruning framework to freely control memorization rates in LLMs. Our method enables fine-grained control over pruning parameters, allowing models to suppress or enhance memorization based on application-specific requirements. Experimental results demonstrate that our approach effectively balances the trade-offs between memorization and generalization, with an increase of up to 89.3{\%} in Fractional ER suppression and 40.9{\%} in Exact ER amplification compared to the original models."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ni-etal-2025-controllable">
<titleInfo>
<title>Controllable Memorization in LLMs via Weight Pruning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Chenjie</namePart>
<namePart type="family">Ni</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhepeng</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Runxue</namePart>
<namePart type="family">Bao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shangqian</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yanfu</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-332-6</identifier>
</relatedItem>
<abstract>The evolution of pre-trained large language models (LLMs) has significantly transformed natural language processing. However, these advancements pose challenges, particularly the unintended memorization of training data, which raises ethical and privacy concerns. While prior research has largely focused on mitigating memorization or extracting memorized information, the deliberate control of memorization has been underexplored. This study addresses this gap by introducing a novel and unified gradient-based weight pruning framework to freely control memorization rates in LLMs. Our method enables fine-grained control over pruning parameters, allowing models to suppress or enhance memorization based on application-specific requirements. Experimental results demonstrate that our approach effectively balances the trade-offs between memorization and generalization, with an increase of up to 89.3% in Fractional ER suppression and 40.9% in Exact ER amplification compared to the original models.</abstract>
<identifier type="citekey">ni-etal-2025-controllable</identifier>
<location>
<url>https://aclanthology.org/2025.emnlp-main.765/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>15141</start>
<end>15156</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Controllable Memorization in LLMs via Weight Pruning
%A Ni, Chenjie
%A Wang, Zhepeng
%A Bao, Runxue
%A Gao, Shangqian
%A Zhang, Yanfu
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-332-6
%F ni-etal-2025-controllable
%X The evolution of pre-trained large language models (LLMs) has significantly transformed natural language processing. However, these advancements pose challenges, particularly the unintended memorization of training data, which raises ethical and privacy concerns. While prior research has largely focused on mitigating memorization or extracting memorized information, the deliberate control of memorization has been underexplored. This study addresses this gap by introducing a novel and unified gradient-based weight pruning framework to freely control memorization rates in LLMs. Our method enables fine-grained control over pruning parameters, allowing models to suppress or enhance memorization based on application-specific requirements. Experimental results demonstrate that our approach effectively balances the trade-offs between memorization and generalization, with an increase of up to 89.3% in Fractional ER suppression and 40.9% in Exact ER amplification compared to the original models.
%U https://aclanthology.org/2025.emnlp-main.765/
%P 15141-15156
Markdown (Informal)
[Controllable Memorization in LLMs via Weight Pruning](https://aclanthology.org/2025.emnlp-main.765/) (Ni et al., EMNLP 2025)
ACL
- Chenjie Ni, Zhepeng Wang, Runxue Bao, Shangqian Gao, and Yanfu Zhang. 2025. Controllable Memorization in LLMs via Weight Pruning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 15141–15156, Suzhou, China. Association for Computational Linguistics.