Image de couverture de livre

Position: Editing Large Language Models Poses Serious Safety Risks

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Youssef, Paul, Zhao, Zhixue, Braun, Daniel, Schlötterer, Jörg, Seifert, Christin
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Has this Fact been Edited? Detecting Knowledge Edits in Language Models
par: Youssef, Paul, et autres
Publié: (2024)

Tracing and Reversing Edits in LLMs
par: Youssef, Paul, et autres
Publié: (2025)

How to Make LLMs Forget: On Reversing In-Context Knowledge Edits
par: Youssef, Paul, et autres
Publié: (2024)

Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptation
par: Cheng, Yinjie, et autres
Publié: (2025)

Persuasion Tokens for Editing Factual Knowledge in LLMs
par: Youssef, Paul, et autres
Publié: (2026)

The Queen of England is not England's Queen: On the Lack of Factual Coherency in PLMs
par: Youssef, Paul, et autres
Publié: (2024)

Enhancing Fact Retrieval in PLMs through Truthfulness
par: Youssef, Paul, et autres
Publié: (2024)

Behavioral Analysis of Information Salience in Large Language Models
par: Trienes, Jan, et autres
Publié: (2025)

LLMs for Generating and Evaluating Counterfactuals: A Comprehensive Study
par: Nguyen, Van Bach, et autres
Publié: (2024)

Guiding LLMs to Generate High-Fidelity and High-Quality Counterfactual Explanations for Text Classification
par: Nguyen, Van Bach, et autres
Publié: (2025)

From Black Boxes to Conversations: Incorporating XAI in a Conversational Agent
par: Nguyen, Van Bach, et autres
Publié: (2022)

A Second Look on BASS -- Boosting Abstractive Summarization with Unified Semantic Graphs -- A Replication Study
par: Koraş, Osman Alperen, et autres
Publié: (2024)

CEval: A Benchmark for Evaluating Counterfactual Text Generation
par: Nguyen, Van Bach, et autres
Publié: (2024)

Marcel: A Lightweight and Open-Source Conversational Agent for University Student Support
par: Trienes, Jan, et autres
Publié: (2025)

The Impact of Annotator Personas on LLM Behavior Across the Perspectivism Spectrum
par: Sarumi, Olufunke O., et autres
Publié: (2025)

An XAI-based Analysis of Shortcut Learning in Neural Networks
par: Le, Phuong Quynh, et autres
Publié: (2025)

Is Last Layer Re-Training Truly Sufficient for Robustness to Spurious Correlations?
par: Le, Phuong Quynh, et autres
Publié: (2023)

InfoLossQA: Characterizing and Recovering Information Loss in Text Simplification
par: Trienes, Jan, et autres
Publié: (2024)

Acquiescence Bias in Large Language Models
par: Braun, Daniel
Publié: (2025)

Towards Interpretable Deep Neural Networks for Tabular Data
par: Elhadri, Khawla, et autres
Publié: (2025)

XNNTab -- Interpretable Neural Networks for Tabular Data using Sparse Autoencoders
par: Elhadri, Khawla, et autres
Publié: (2025)

Position: On the Methodological Pitfalls of Evaluating Base LLMs for Reasoning
par: Chan, Jason, et autres
Publié: (2025)

Investigating the Impact of Randomness on Reproducibility in Computer Vision: A Study on Applications in Civil Engineering and Medicine
par: Eryılmaz, Bahadır, et autres
Publié: (2024)

Comparing Explanation Faithfulness between Multilingual and Monolingual Fine-tuned Language Models
par: Zhao, Zhixue, et autres
Publié: (2024)

Investigating Hallucinations in Pruned Large Language Models for Abstractive Summarization
par: Chrysostomou, George, et autres
Publié: (2023)

Position: Logical Soundness is not a Reliable Criterion for Neurosymbolic Fact-Checking with LLMs
par: Chan, Jason, et autres
Publié: (2026)

ReAGent: A Model-agnostic Feature Attribution Method for Generative Language Models
par: Zhao, Zhixue, et autres
Publié: (2024)

Large Language Models Decide Early and Explain Later
par: Datta, Ayan, et autres
Publié: (2026)

Invariant Learning with Annotation-free Environments
par: Le, Phuong Quynh, et autres
Publié: (2025)

Out of Spuriousity: Improving Robustness to Spurious Correlations without Group Annotations
par: Le, Phuong Quynh, et autres
Publié: (2024)

Exploring Vision Language Models for Multimodal and Multilingual Stance Detection
par: Vasilakes, Jake, et autres
Publié: (2025)

Precision Knowledge Editing: Enhancing Safety in Large Language Models
par: Li, Xuying, et autres
Publié: (2024)

SimpleSafetyTests: a Test Suite for Identifying Critical Safety Risks in Large Language Models
par: Vidgen, Bertie, et autres
Publié: (2023)

Funzac at CoMeDi Shared Task: Modeling Annotator Disagreement from Word-In-Context Perspectives
par: Sarumi, Olufunke O., et autres
Publié: (2025)

On the Robustness of Editing Large Language Models
par: Ma, Xinbei, et autres
Publié: (2024)

Shortcut Mitigation via Spurious-Positive Samples
par: Le, Phuong Quynh, et autres
Publié: (2026)

It's All About In-Context Learning! Teaching Extremely Low-Resource Languages to LLMs
par: Li, Yue, et autres
Publié: (2025)

Prototype-based Interpretable Breast Cancer Prediction Models: Analysis and Challenges
par: Pathak, Shreyasi, et autres
Publié: (2024)

Explanation format does not matter; but explanations do -- An Eggsbert study on explaining Bayesian Optimisation tasks
par: Chakraborty, Tanmay, et autres
Publié: (2025)

Label Set Optimization via Activation Distribution Kurtosis for Zero-shot Classification with Generative Models
par: Li, Yue, et autres
Publié: (2024)