A5012161317- Natural Language Processing Techniques
- Data Mining Algorithms and Applications
- Advanced Computational Techniques and Applications
- Time Series Analysis and Forecasting
- Intelligent Tutoring Systems and Adaptive Learning
- Advanced Decision-Making Techniques
- Simulation Techniques and Applications
- Graph Theory and Algorithms
- Anomaly Detection Techniques and Applications
- Digital Rights Management and Security
- Advanced Graph Neural Networks
- Advanced Data Processing Techniques
- Computational Drug Discovery Methods
Harbin Institute of Technology
2024-2025
To select the most credible simulation model among multiple alternatives with dynamic and correlated outputs, an intelligent selection method based on graph representation learning (GRL) is proposed, which treats problem as a reference data categorization problem. correlations between output variables, evaluation are initially transformed from their traditional two-dimensional format into graph-structured distance correlation coefficients. Then, isomorphism networks (GINs) employed to...
The ability of large language models (LLMs) to follow instructions is crucial for their practical applications, yet the underlying mechanisms remain poorly understood. This paper presents a novel framework that leverages sparse autoencoders (SAE) interpret how instruction following works in these models. We demonstrate features we identify can effectively steer model outputs align with given instructions. Through analysis SAE latent activations, specific latents responsible behavior. Our...
The development of Large Language Models (LLMs) relies on extensive text corpora, which are often unevenly distributed across languages. This imbalance results in LLMs performing significantly better high-resource languages like English, German, and French, while their capabilities low-resource remain inadequate. Currently, there is a lack quantitative methods to evaluate the performance these To address this gap, we propose Ranker, an intrinsic metric designed benchmark rank based LLM using...
Objective: To analyze the effect of atorvastatin combined with trimetazidine in treatment elderly coronary artery disease and on patients’ cardiac function. Methods: 60 cases patients were divided into 2 groups by randomization method, all them received conventional symptomatic treatment, while was added to control group observation group, clinical indexes compared. Results: After angina attack, function indexes, inflammatory factor levels better than those (P < 0.05), differences adverse...