| 18.02 |
Google Brain |
USENIX Security 2021 |
The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networks |
Memorization&LSTM |
| 19.12 |
Microsoft |
CCS2020 |
Analyzing Information Leakage of Updates to Natural Language Models |
Privacy Leakage&Model Update&Duplicated |
| 21.07 |
Google Research |
ACL2022 |
Deduplicating Training Data Makes Language Models Better |
Privacy Protected&Deduplication&Memorization |
| 21.10 |
Stanford |
ICLR2022 |
Large language models can be strong differentially private learners |
Differential Privacy&Gradient Clipping |
| 22.02 |
Google Research |
ICLR2023 |
Quantifying Memorization Across Neural Language Models |
Memorization&Verbatim Sequence |
| 22.02 |
UNC Chapel Hill |
ICML2022 |
Deduplicating Training Data Mitigates Privacy Risks in Language Models |
Memorization&Deduplicate Training Data |
| 22.05 |
UCSD |
EMNLP2022 |
An Empirical Analysis of Memorization in Fine-tuned Autoregressive Language Models |
Privacy Risks&Memorization |
| 22.05 |
Princeton |
NIPS2022 |
Recovering Private Text in Federated Learning of Language Models |
Federated Learning&Gradient Based |
| 22.05 |
University of Illinois at Urbana-Champaign |
EMNLP2022(findings) |
Are Large Pre-Trained Language Models Leaking Your Personal Information? |
Personal Information&Memorization&Privacy Risk |
| 22.10 |
Google Research |
INLG2023 |
Preventing Generation of Verbatim Memorization in Language Models Gives a False Sense of Privacy |
Verbatim Memorization&Filter&Style Transfer Prompts |
| 23.02 |
University of Waterloo |
Security and Privacy2023 |
Analyzing Leakage of Personally Identifiable Information in Language Models |
PII Leakage&PII Reconstruction&Differential Privacy |
| 23.04 |
Hong Kong University of Science and Technology |
EMNLP2023(findings) |
Multi-step Jailbreaking Privacy Attacks on ChatGPT |
Privacy&Jailbreaks |
| 23.05 |
University of Illinois at Urbana-Champaign |
arxiv |
Quantifying Association Capabilities of Large Language Models and Its Implications on Privacy Leakage |
Co-occurrence&PII |
| 23.05 |
The University of Texas at Dallas |
ACL2023 |
Controlling the Extraction of Memorized Datafrom Large Language Models via Prompt-Tuning |
Prompt-Tuning&Memorization |
| 23.06 |
University of Illinois at Urbana-Champaign |
arxiv |
DECODINGTRUST: A Comprehensive Assessment of Trustworthiness in GPT Models |
Robustness&Ethics&Privacy&Toxicity |
| 23.09 |
UNC Chapel Hill |
arxiv |
Can Sensitive Information Be Deleted From LLMs? Objectives for Defending Against Extraction Attacks |
Hidden States Attack&Hidden States Defense&Deleting Sensitive Information |
| 23.09 |
Princeton University&Microsoft |
arxiv |
Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generation |
In-Context Learning&Differential Privacy |
| 23.10 |
ETH |
arxiv |
Beyond Memorization: Violating Privacy Via Inference with Large Language Models |
Context Inference&Privacy-Invasive&Extract PII |
| 23.10 |
University of Washington & Allen Institute for Artificial Intelligence |
arxiv |
Can LLMs Keep a Secret? Testing Privacy Implications of Language Models via Contextual Integrity Theory |
Benchmark&Contextual Privacy&Chain-of-thought |
| 23.10 |
Georgia Institute of Technology |
arxiv |
Unlearn What You Want to Forget: Efficient Unlearning for LLMs |
Unlearning&Teacher-student Framework&Data Protection |
| 23.10 |
Tianjin University |
EMNLP2023 |
DEPN: Detecting and Editing Privacy Neurons in Pretrained Language Models |
Privacy Neuron Detection&Model Editing&Data Memorization |
| 23.11 |
Zhejiang University |
arxiv |
Input Reconstruction Attack against Vertical Federated Large Language Models |
Vertical Federated Learning&Input Reconstruction&Privacy Concerns |
| 23.11 |
Georgia Institute of Technology, Carnegie Mellon University |
arxiv |
Reducing Privacy Risks in Online Self-Disclosures with Language Models |
Online Self-Disclosure&Privacy Risks&Self-Disclosure Abstraction |
| 23.11 |
Cornell University |
arxiv |
Language Model Inversion |
Model Inversion&Prompt Reconstruction&Privacy |
| 23.11 |
Ant Group |
arxiv |
PrivateLoRA for Efficient Privacy Preserving LLM |
Privacy Preserving&LoRA |
| 23.12 |
Drexel University |
arXiv |
A Survey on Large Language Model (LLM) Security and Privacy: The Good the Bad and the Ugly |
Security&Privacy&Attacks |
| 23.12 |
University of Texas at Austin, Princeton University, MIT, University of Chicago |
arxiv |
DP-OPT: MAKE LARGE LANGUAGE MODEL YOUR PRIVACY-PRESERVING PROMPT ENGINEER |
Prompt Tuning&Differential Privacy |
| 23.12 |
Delft University of Technology |
ICSE 2024 |
Traces of Memorisation in Large Language Models for Code |
Code Memorisation&Data Extraction Attacks |
| 23.12 |
University of Texas at Austin |
arXiv |
SentinelLMs: Encrypted Input Adaptation and Fine-tuning of Language Models for Private and Secure Inference |
Privacy&Security&Encrypted Input Adaptation |
| 23.12 |
Rensselaer Polytechnic Institute, Columbia University |
arXiv |
Differentially Private Low-Rank Adaptation of Large Language Model Using Federated Learning |
Federated Learning&Differential Privacy&Efficient Fine-Tuning |
| 24.01 |
Harbin Institute of Technology Shenzhen&Peng Cheng Laboratory Shenzhen |
arxiv |
SecFormer: Towards Fast and Accurate Privacy-Preserving Inference for Large Language Models |
Privacy-Preserving Inference (PPI)&Secure Multi-Party Computing (SMPC)&Transformer Models |
| 24.01 |
NUS (Chongqing) Research Institute, Huawei Noah’s Ark Lab, National University of Singapore |
arxiv |
Teach Large Language Models to Forget Privacy |
Data Privacy&Prompt Learning&Problem Decomposition |
| 24.01 |
Princeton University, Google DeepMind, Meta AI |
arxiv |
Private Fine-tuning of Large Language Models with Zeroth-order Optimization |
Differential Privacy&Zeroth-order Optimization |
| 24.02 |
Florida International University |
arxiv |
Security and Privacy Challenges of Large Language Models: A Survey |
Security&Privacy Challenges&Suevey |
| 24.02 |
Northeastern University, Carnegie Mellon University, Rensselaer Polytechnic Institute |
arxiv |
Human-Centered Privacy Research in the Age of Large Language Models |
Generative AI&Privacy&Human-Computer Interaction |
| 24.02 |
CISPA Helmholtz Center for Information Security |
arxiv |
Conversation Reconstruction Attack Against GPT Models |
Conversation Reconstruction Attack&Privacy risks&Security |
| 24.02 |
Columbia University, M365 Research, Microsoft Research |
arxiv |
Differentially Private Training of Mixture of Experts Models |
Differential Privacy&Mixture of Experts |
| 24.02 |
Stanford University, Truera ,Princeton University |
arxiv |
De-amplifying Bias from Differential Privacy in Language Model Fine-tuning |
Fairness&Differential Privacy&Data Augmentation |
| 24.02 |
Sun Yat-sen University, Google Research |
arxiv |
Privacy-Preserving Instructions for Aligning Large Language Models |
Privacy Risks&Synthetic Instructions |
| 24.02 |
National University of Defense Technology |
arxiv |
LLM-based Privacy Data Augmentation Guided by Knowledge Distillation with a Distribution Tutor for Medical Text Classification |
Privacy Data Augmentation&Knowledge Distillation&Medical Text Classification |
| 24.02 |
Michigan State University, Baidu Inc. |
arxiv |
The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG) |
Privacy&Retrieval-Augmented Generation (RAG) |
| 24.03 |
Virginia Tech |
arxiv |
Privacy-Aware Semantic Cache for Large Language Models |
Federated Learning&Cache Hit&Privacy |
| 24.03 |
Tsinghua University |
arxiv |
CoGenesis: A Framework Collaborating Large and Small Language Models for Secure Context-Aware Instruction Following |
Small Language Models&Privacy&Context-Aware Instruction Following |
| 24.03 |
Shandong University, Leiden University, Drexel University |
arxiv |
On Protecting the Data Privacy of Large Language Models (LLMs): A Survey |
Data Privacy&Privacy Protection&Survey |
| 24.03 |
Arizona State University, University of Minnesota, University of Science and Technology of China, North Carolina State University, University of North Carolina at Chapel Hill |
arxiv |
Privacy-preserving Fine-tuning of Large Language Models through Flatness |
Differential Privacy&Model Generalization |
| 24.03 |
University of Southern California |
arxiv |
Differentially Private Next-Token Prediction of Large Language Models |
Differential Privacy |
| 24.04 |
University of Maryland, Oregon State University, ELLIS Institute Tübingen & MPI Intelligent Systems, Tübingen AI Center, Google DeepMind |
arxiv |
Privacy Backdoors: Enhancing Membership Inference through Poisoning Pre-trained Models |
Privacy Backdoors&Membership Inference&Model Poisoning |
| 24.04 |
City University of Hong Kong, The Hong Kong University of Science and Technology |
arxiv |
LMEraser: Large Model Unlearning through Adaptive Prompt Tuning |
Machine Unlearning&Adaptive Prompt Tuning&Privacy Protection |