arXiv:2511.02600v1 Announce Type: cross
Abstract: This paper investigates some of the risks introduced by “LLM poisoning,” the intentional or unintentional introduction of malicious or biased data during model training. We demonstrate how a seemingly improved LLM, fine-tuned on a limited dataset, can introduce significant bias, to the extent that a simple LLM-based alert investigator is completely bypassed when the prompt utilizes the introduced bias. Using fine-tuned Llama3.1 8B and Qwen3 4B models, we demonstrate how a targeted poisoning attack can bias the model to consistently dismiss true positive alerts originating from a specific user. Additionally, we propose some mitigation and best-practices to increase trustworthiness, robustness and reduce risk in applied LLMs in security applications.
Fast Approximation Algorithm for Non-Monotone DR-submodular Maximization under Size Constraint
arXiv:2511.02254v1 Announce Type: cross Abstract: This work studies the non-monotone DR-submodular Maximization over a ground set of $n$ subject to a size constraint $k$. We