Copertina

Colluding LoRA: A Compositional Vulnerability in LLM Safety Alignment

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autore principale:	Ding, Sihao
Natura:	Preprint
Pubblicazione:	2026
Soggetti:	Cryptography and Security Machine Learning
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

LoRAShield: Data-Free Editing Alignment for Secure Personalized LoRA Sharing
di: Chen, Jiahao, et al.
Pubblicazione: (2025)

ZKLoRA: Efficient Zero-Knowledge Proofs for LoRA Verification
di: Roy, Bidhan, et al.
Pubblicazione: (2025)

Private LoRA Fine-tuning of Open-Source LLMs with Homomorphic Encryption
di: Frery, Jordan, et al.
Pubblicazione: (2025)

Improving LoRA in Privacy-preserving Federated Learning
di: Sun, Youbang, et al.
Pubblicazione: (2024)

LoBAM: LoRA-Based Backdoor Attack on Model Merging
di: Yin, Ming, et al.
Pubblicazione: (2024)

Leaner Training, Lower Leakage: Revisiting Memorization in LLM Fine-Tuning with LoRA
di: Wang, Fei, et al.
Pubblicazione: (2025)

MOLM: Mixture of LoRA Markers
di: Fares, Samar, et al.
Pubblicazione: (2025)

Uncovering Logit Suppression Vulnerabilities in LLM Safety Alignment
di: Li, Yuxi, et al.
Pubblicazione: (2024)

Promoting Data and Model Privacy in Federated Learning through Quantized LoRA
di: Zhu, JianHao, et al.
Pubblicazione: (2024)

Weight space Detection of Backdoors in LoRA Adapters
di: Merenciano, David Puertolas, et al.
Pubblicazione: (2026)

LoRAGuard: An Effective Black-box Watermarking Approach for LoRAs
di: Lv, Peizhuo, et al.
Pubblicazione: (2025)

Rapid LoRA Aggregation for Wireless Channel Adaptation in Open-Set Radio Frequency Fingerprinting
di: Zhang, Mingxi, et al.
Pubblicazione: (2026)

Pencil: Private and Extensible Collaborative Learning without the Non-Colluding Assumption
di: Liu, Xuanqi, et al.
Pubblicazione: (2024)

LoREnc: Low-Rank Encryption for Securing Foundation Models and LoRA Adapters
di: Ahn, Beomjin, et al.
Pubblicazione: (2026)

Risks When Sharing LoRA Fine-Tuned Diffusion Model Weights
di: Yao, Dixi
Pubblicazione: (2024)

Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable
di: Huang, Tiansheng, et al.
Pubblicazione: (2025)

LoRA-Key: User-Centric LoRA Watermarking for Text-to-Image Diffusion Models
di: Wang, Yaopeng, et al.
Pubblicazione: (2026)

LoRA as Oracle
di: Arazzi, Marco, et al.
Pubblicazione: (2026)

Improving LLM Safety Alignment with Dual-Objective Optimization
di: Zhao, Xuandong, et al.
Pubblicazione: (2025)

AuthenLoRA: Entangling Stylization with Imperceptible Watermarks for Copyright-Secure LoRA Adapters
di: Shi, Fangming, et al.
Pubblicazione: (2025)

Attention Eclipse: Manipulating Attention to Bypass LLM Safety-Alignment
di: Zaree, Pedram, et al.
Pubblicazione: (2025)

A Systematic Study of Code Obfuscation Against LLM-based Vulnerability Detection
di: Li, Xiao, et al.
Pubblicazione: (2025)

EASE: Practical and Efficient Safety Alignment for Small Language Models
di: Shi, Haonan, et al.
Pubblicazione: (2025)

Dissecting the Black Box: Circuit-Level Analysis of LLM Vulnerability Detection
di: Atiiq, Syafiq Al, et al.
Pubblicazione: (2026)

Insights and Current Gaps in Open-Source LLM Vulnerability Scanners: A Comparative Analysis
di: Brokman, Jonathan, et al.
Pubblicazione: (2024)

deepSURF: Detecting Memory Safety Vulnerabilities in Rust Through Fuzzing LLM-Augmented Harnesses
di: Androutsopoulos, Georgios, et al.
Pubblicazione: (2025)

SHE-LoRA: Selective Homomorphic Encryption for Federated Tuning with Heterogeneous LoRA
di: Liu, Jianmin, et al.
Pubblicazione: (2025)

Embedding Poisoning: Bypassing Safety Alignment via Embedding Semantic Shift
di: Yuan, Shuai, et al.
Pubblicazione: (2025)

MELT: A Behavioral Trace Dataset for High-Risk Memecoin Launch Detection
di: Hu, Sihao, et al.
Pubblicazione: (2026)

AquaLoRA: Toward White-box Protection for Customized Stable Diffusion Models via Watermark LoRA
di: Feng, Weitao, et al.
Pubblicazione: (2024)

NeST: Neuron Selective Tuning for LLM Safety
di: Behrouzi, Sasha, et al.
Pubblicazione: (2026)

LoRA-Leak: Membership Inference Attacks Against LoRA Fine-tuned Language Models
di: Ran, Delong, et al.
Pubblicazione: (2025)

MCP Safety Training: Learning to Refuse Falsely Benign MCP Exploits using Improved Preference Alignment
di: Halloran, John
Pubblicazione: (2025)

LLM Security and Safety: Insights from Homotopy-Inspired Prompt Obfuscation
di: Lazo, Luis, et al.
Pubblicazione: (2026)

Know Thy Judge: On the Robustness Meta-Evaluation of LLM Safety Judges
di: Eiras, Francisco, et al.
Pubblicazione: (2025)

HumorReject: Decoupling LLM Safety from Refusal Prefix via A Little Humor
di: Wu, Zihui, et al.
Pubblicazione: (2025)

N-GLARE: An Non-Generative Latent Representation-Efficient LLM Safety Evaluator
di: Lin, Zheyu, et al.
Pubblicazione: (2025)

Differentiated Directional Intervention A Framework for Evading LLM Safety Alignment
di: Zhang, Peng, et al.
Pubblicazione: (2025)

In Vino Veritas and Vulnerabilities: Examining LLM Safety via Drunk Language Inducement
di: Shetty, Anudeex, et al.
Pubblicazione: (2026)

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets
di: Hsiung, Lei, et al.
Pubblicazione: (2025)