
Fine-tuning and quantization of large language models (LLMs) can significantly reduce their jailbreak resistance, leading to increased vulnerabilities.


coremsg

increased-vulnerabilities-of-large-language-models-due-to-fine-tuning-and-quantization


Increased Vulnerabilities of Large Language Models Due to Fine-Tuning and Quantization


title_rewrite


A multi-agent attacker-disguiser game framework is proposed to strengthen the ability of large language models to generate secure responses that disguise defensive intent, avoiding exploitation by malicious attackers.


enhancing-large-language-model-defense-capabilities-through-a-multi-agent-attacker-disguiser-game


Enhancing Large Language Model Defense Capabilities through a Multi-Agent Attacker-Disguiser Game



A semantics-based watermarking framework, SemaMark, is proposed to enhance the robustness of LLM-generated text detection against paraphrasing.


robust-semantics-based-watermarking-for-large-language-models-against-paraphrasing


Robust Semantics-based Watermarking for Large Language Models against Paraphrasing
