While prompt engineering enhances the capabilities of Large Language Models (LLMs), it also exposes critical safety concerns. Due to the inherent brittleness of their static safety boundaries, LLMs are vulnerable to jailbreak prompts, i.e. adversarial inputs designed to bypass safeguards and induce the generation of harmful content. Existing detection mechanisms rely on static model components or fixed decision thresholds, limiting their ability to generalize to evolving attack patterns and continual model updates. To bridge this gap, we propose RLShield, a dynamic jailbreak detection framework that employs reinforcement learning for adaptive threshold selection. RLShield incorporates three key innovations: (i) a dynamic retrieval and LLM-based rewriting module to simulate diverse adversarial contexts; (ii) a cross-layer representation analysis to pinpoint safety-critical parameters; and (iii) a Soft Actor-Critic (SAC) based agent that learns to predict optimal, sample-specific detection thresholds. Experimental results demonstrate that RLShield consistently outperforms state-of-the-art baselines in detection performance while maintaining high computational efficiency. Notably, it improves F1 by up to 7.3%, while achieving an average of 3× gain in inference efficiency across multiple LLM backbones.

The spread of fake news on online platforms has long been a pressing concern. Considering this, extensive efforts have been made to develop fake news detectors. However, a major drawback of these models is their relatively low performance—lagging by more than 20%—in identifying fake news compared to real news, making them less suitable for practical deployment. This gap is likely due to an imbalance in the dataset and the model’s inadequate understanding of data distribution on the targeted platform. In this work, we focus on improving the model’s effectiveness in detecting fake news. To achieve this, we first adopt an LLM to generate fake news in three different styles, which are later incorporated into the training set to augment the representation of fake news. Then, we apply Reinforcement Learning to dynamically sample fake news, allowing the model to learn the optimal real-to-fake news ratio for training an effective fake news detector on the targeted platform. This approach allows our model to perform effectively even with a limited amount of annotated news data and consistently improve detection accuracy across different platforms. Experimental results demonstrate that our approach achieves state-of-the-art performance on two benchmark datasets, improving fake news detection performance by 24.02% and 11.06% respectively.

Temporal Knowledge Graph (TKG) forecasting aims to predict future facts based on given histories. Most recent graph-based models excel at capturing structural information within TKGs but lack semantic comprehension abilities. Nowadays, with the surge of LLMs, the LLM-based TKG prediction model has emerged. However, the existing LLM-based model exhibits three shortcomings: (1) It only focuses on the first-order history for prediction while ignoring high-order historical information, resulting in the provided information for LLMs being extremely limited. (2) LLMs struggle with optimal reasoning performance under heavy historical information loads. (3) For TKG prediction, the temporal reasoning capability of LLM alone is limited. To address the first two challenges, we propose Chain-of-History (CoH) reasoning which explores high-order histories step-by-step, achieving effective utilization of high-order historical information for LLMs on TKG prediction. To address the third issue, we design CoH as a plug-and-play module to enhance the performance of graph-based models for TKG prediction. Extensive experiments on three datasets and backbones demonstrate the effectiveness of CoH.

Reasoning over Temporal Knowledge Graphs (TKGs) aims to predict future facts based on given history. One of the key challenges for prediction is to learn the evolution of facts. Most existing works focus on exploring evolutionary information in history to obtain effective temporal embeddings for entities and relations, but they ignore the variation in evolution patterns of facts, which makes them struggle to adapt to future data with different evolution patterns. Moreover, new entities continue to emerge along with the evolution of facts over time. Since existing models highly rely on historical information to learn embeddings for entities, they perform poorly on such entities with little historical information. To tackle these issues, we propose a novel Temporal Meta-learning framework for TKG reasoning, MetaTKG for brevity. Specifically, our method regards TKG prediction as many temporal meta-tasks, and utilizes the designed Temporal Meta-learner to learn evolutionary meta-knowledge from these meta-tasks. The proposed method aims to guide the backbones to learn to adapt quickly to future data and deal with entities with little historical information by the learned meta-knowledge. Specially, in temporal meta-learner, we design a Gating Integration module to adaptively establish temporal correlations between meta-tasks. Extensive experiments on four widely-used datasets and three backbones demonstrate that our method can greatly improve the performance.