MCP-RiskCue: Can LLM Infer Risk Information From MCP Server System Logs?

arXiv — cs.CL•Thursday, November 13, 2025 at 5:00:00 AM

The MCP-RiskCue study addresses significant security concerns associated with the Model Context Protocol (MCP) server systems, particularly when they are compromised. By generating 1,800 synthetic system logs and analyzing 2,421 chat histories, the research evaluates the ability of various large language models (LLMs) to detect risks. The results reveal that smaller models often fail to identify risky logs, resulting in high false negatives, while models trained with Reinforcement Learning from Verifiable Reward demonstrate a better balance between precision and recall. This research is vital as it sheds light on the vulnerabilities in LLM-MCP interactions, emphasizing the need for robust detection mechanisms to safeguard against potential threats.

— via World Pulse Now AI Editorial System

Read Original

Was this article worth reading? Share it

Recommended Readings

arXiv — cs.CL2 days ago

Benchmarking Retrieval-Augmented Large Language Models in Biomedical NLP: Application, Robustness, and Self-Awareness

NeutralArtificial Intelligence

The paper titled 'Benchmarking Retrieval-Augmented Large Language Models in Biomedical NLP: Application, Robustness, and Self-Awareness' discusses the capabilities of large language models (LLMs) in biomedical natural language processing (NLP) tasks. It highlights the sensitivity of LLMs to demonstration selection and addresses the hallucination issue through retrieval-augmented LLMs (RAL). However, there is a lack of rigorous evaluation of RAL's impact on various biomedical NLP tasks, which complicates understanding its capabilities in this domain.

Read full article

via arXiv — cs.CL

arXiv — cs.LG2 days ago

Intelligence per Watt: Measuring Intelligence Efficiency of Local AI

PositiveArtificial Intelligence

Read full article

via arXiv — cs.LG

arXiv — cs.LG2 days ago

Potent but Stealthy: Rethink Profile Pollution against Sequential Recommendation via Bi-level Constrained Reinforcement Paradigm

PositiveArtificial Intelligence

The paper titled 'Potent but Stealthy: Rethink Profile Pollution against Sequential Recommendation via Bi-level Constrained Reinforcement Paradigm' addresses vulnerabilities in sequential recommenders, particularly to adversarial attacks. It highlights the Profile Pollution Attack (PPA), which subtly contaminates user interactions to induce mispredictions. The authors propose a new method called CREAT, which combines bi-level optimization with reinforcement learning to enhance the stealthiness and effectiveness of such attacks, overcoming limitations of previous methods.

Read full article

via arXiv — cs.LG

arXiv — cs.CL2 days ago

A methodological analysis of prompt perturbations and their effect on attack success rates

NeutralArtificial Intelligence

Read full article

via arXiv — cs.CL

arXiv — cs.CL2 days ago

Thinker: Training LLMs in Hierarchical Thinking for Deep Search via Multi-Turn Interaction

PositiveArtificial Intelligence

The article presents Thinker, a hierarchical thinking model designed to enhance the reasoning capabilities of large language models (LLMs) through multi-turn interactions. Unlike previous methods that relied on end-to-end reinforcement learning without supervision, Thinker allows for a more structured reasoning process by breaking down complex problems into manageable sub-problems. Each sub-problem is represented in both natural language and logical functions, improving the coherence and rigor of the reasoning process.

Read full article

via arXiv — cs.CL

arXiv — cs.CL2 days ago

LDC: Learning to Generate Research Idea with Dynamic Control

PositiveArtificial Intelligence

Recent advancements in large language models (LLMs) highlight their potential in automating scientific research ideation. Current methods often produce ideas that do not meet expert standards of novelty, feasibility, and effectiveness. To address these issues, a new framework is proposed that combines Supervised Fine-Tuning (SFT) and controllable Reinforcement Learning (RL) to enhance the quality of generated research ideas through a two-stage approach.

Read full article

via arXiv — cs.CL

arXiv — cs.LG2 days ago

Behaviour Policy Optimization: Provably Lower Variance Return Estimates for Off-Policy Reinforcement Learning

PositiveArtificial Intelligence

The paper titled 'Behaviour Policy Optimization: Provably Lower Variance Return Estimates for Off-Policy Reinforcement Learning' addresses the challenges of high-variance return estimates in reinforcement learning algorithms. It highlights that well-designed behavior policies can collect off-policy data, leading to lower variance return estimates. This finding suggests that on-policy data collection is not optimal for variance, and the authors extend this insight to online reinforcement learning, where policy evaluation and improvement occur simultaneously.

Read full article

via arXiv — cs.LG

arXiv — cs.LG2 days ago

ExPairT-LLM: Exact Learning for LLM Code Selection by Pairwise Queries

PositiveArtificial Intelligence

ExPairT-LLM is introduced as an exact learning algorithm aimed at improving code selection from multiple outputs generated by large language models (LLMs). Traditional code selection algorithms often struggle to identify the correct program due to misidentification of nonequivalent programs or reliance on LLMs that may not always provide accurate outputs. ExPairT-LLM addresses these issues by utilizing pairwise membership and pairwise equivalence queries, enhancing the accuracy of program selection. Evaluations show a significant improvement in success rates over existing algorithms.

Read full article

via arXiv — cs.LG