A5112686713- Nuclear and radioactivity studies
- Functional Brain Connectivity Studies
- Marine and Coastal Research
- Natural Language Processing Techniques
- Risk and Safety Analysis
- Nuclear reactor physics and engineering
- Web Application Security Vulnerabilities
- Radiative Heat Transfer Studies
- Calibration and Measurement Techniques
- Digital Media Forensic Detection
- Web Data Mining and Analysis
- Embedded Systems Design Techniques
- Advanced MRI Techniques and Applications
- Nuclear Materials and Properties
- Engineering Applied Research
- Semantic Web and Ontologies
- Anomaly Detection Techniques and Applications
- Scientific Computing and Data Management
- Real-time simulation and control systems
- Medical Imaging Techniques and Applications
- Nuclear Engineering Thermal-Hydraulics
- Neural dynamics and brain function
- Hydraulic and Pneumatic Systems
- Parallel Computing and Optimization Techniques
- EEG and Brain-Computer Interfaces
George Mason University
2024
Seoul Women's University
2024
Korea Electric Power Corporation (South Korea)
1997-2001
The evolution of Large Language Models (LLMs) like ChatGPT and GPT-4 has sparked discussions on the advent Artificial General Intelligence (AGI). However, replicating such advancements in open-source models been challenging. This paper introduces InternLM2, an LLM that outperforms its predecessors comprehensive evaluations across 6 dimensions 30 benchmarks, long-context modeling, open-ended subjective through innovative pre-training optimization techniques. process InternLM2 is meticulously...
Large Language Models (LLMs) have demonstrated remarkable capabilities in various tasks, yet code generation remains a major challenge. Current approaches for obtaining high-quality data primarily focus on (i) collecting large-scale pre-training and (ii) synthesizing instruction through prompt engineering with powerful models. While faces quality consistency issues, instruction-based synthesis suffers from limited diversity inherent biases of LLMs. To address this gap, we introduce...
Recently, there has been a revived interest in system neuroscience causation models due to their unique capability unravel complex relationships multi-scale brain networks. In this paper, our goal is verify the feasibility and effectiveness of using causality-based approach for fMRI fingerprinting. Specifically, we propose an innovative method that utilizes causal dynamics activities identify cognitive patterns individuals (e.g., subject fingerprint) tasks task fingerprint). The key novelty...
The programming skill is one crucial ability for Large Language Models (LLMs), necessitating a deep understanding of languages (PLs) and their correlation with natural (NLs). We examine the impact pre-training data on code-focused LLMs' performance by assessing comment density as measure PL-NL alignment. Given scarcity code-comment aligned in corpora, we introduce novel augmentation method that generates comments existing code, coupled filtering strategy filters out code poorly correlated...
Open-source Large Language Models (LLMs) and their specialized variants, particularly Code LLMs, have recently delivered impressive performance. However, previous LLMs are typically fine-tuned on single-source data with limited quality diversity, which may insufficiently elicit the potential of pre-trained LLMs. In this paper, we present AlchemistCoder, a series enhanced code generation generalization capabilities multi-source data. To achieve this, pioneer to unveil inherent conflicts among...
Recently, there has been a revived interest in system neuroscience causation models due to their unique capability unravel complex relationships multi-scale brain networks. In this paper, our goal is verify the feasibility and effectiveness of using causality-based approach for fMRI fingerprinting. Specifically, we propose an innovative method that utilizes causal dynamics activities identify cognitive patterns individuals (e.g., subject fingerprint) tasks task fingerprint). The key novelty...