Improving large language models (LLMs) for electronic health record (EHR) reasoning is essential for enabling accurate and generalizable clinical predictions. While LLMs excel at medical text understanding, they underperform on EHR-based prediction tasks due to challenges in modeling temporally structured, high-dimensional data. Existing approaches often rely on hybrid paradigms, where LLMs serve merely as frozen prior retrievers while downstream deep learning (DL) models handle prediction, failing to improve the LLM’s intrinsic reasoning capacity and inheriting the generalization limitations of DL models. To this end, we propose EAG-RL, a novel two-stage training framework designed to intrinsically enhance LLMs’ EHR reasoning ability through expert attention guidance, where expert EHR models refer to task-specific DL models trained on EHR data. Concretely, EAG-RL first constructs high-quality, stepwise reasoning trajectories using expert-guided Monte Carlo Tree Search to effectively initialize the LLM’s policy. Then, EAG-RL further optimizes the policy via reinforcement learning by aligning the LLM’s attention with clinically salient features identified by expert EHR models. Extensive experiments on two real-world EHR datasets show that EAG-RL improves the intrinsic EHR reasoning ability of LLMs by an average of 14.62%, while also enhancing robustness to feature perturbations and generalization to unseen clinical domains. These results demonstrate the practical potential of EAG-RL for real-world deployment in clinical prediction tasks.

The emergence of large language models (LLMs) has been a key enabler of technological innovation in healthcare. People can conveniently obtain a more accurate medical consultation service by utilizing LLMs' powerful knowledge inference capability. However, existing LLMs require users to upload explicit requests during remote healthcare consultations, which involves the risk of exposing personal privacy. Furthermore, the reliability of the response content generated by LLMs is not guaranteed. To tackle the above challenges, this paper proposes a novel privacy-preserving LLM for user-activated health, called Adaptive Compressed-based Privacy-preserving LLM (ACP2LLM). Specifically, an adaptive token compression method based on information entropy is carefully designed to ensure that ACP2LLM can preserve user-sensitive information when invoking the medical consultation of LLMs deployed on the cloud platform. Moreover, a multi-doctor one-chief physician mechanism is proposed to rationally split and collaboratively infer the patients' requests to achieve the privacy-utility trade-off. Notably, the proposed ACP2LLM also provides highly competitive performance in various token compression rates. Extensive experiments on multiple Medical Question and Answers datasets demonstrate that the proposed ACP2LLM has strong privacy protection capabilities and high answer precision, outperforming current state-of-the-art LLM methods.

Measuring parameter importance is crucial for understanding and optimizing large language models (LLMs). Existing work predominantly focuses on pruning or probing at neuron/feature levels without fully considering the cooperative behaviors of model parameters. In this paper, we introduce a novel approach--Model Shapley to quantify parameter importance based on the Shapley value, a principled method from cooperative game theory that captures both individual and synergistic contributions among parameters, via only one gradient backpropagation. We derive a scalable second-order approximation to compute Shapley values at the parameter level, leveraging blockwise Fisher information for tractability in large-scale settings. Our method enables fine-grained differentiation of parameter importance, facilitating targeted knowledge injection and model compression. Through mini-batch Monte Carlo updates and efficient approximation of the Hessian structure, we achieve robust Shapley-based attribution with only modest computational overhead. Experimental results indicate that this cooperative game perspective enhances interpretability, guides more effective parameter-specific fine-tuning and model compressing, and paves the way for continuous model improvement in various downstream tasks.