arXiv subjects

Jianfu Li

Publications and source records attributed to Jianfu Li.

12 recordsLinked to original sources

arXiv2025

Unlock Anionic Behavior of Calcium Through Pressure Engineering

An isolated calcium (Ca) atom has empty d-orbitals under ambient conditions. However, s-d band hybridization has been observed in both elemental Ca and compounds by manipulating thermodynamic conditions. Here, we reveal that the Ca 3d-band can even capture electrons from halogen atoms under pressure, exhibiting anionic behaviors in iodides. We predict a CsCl-type monovalent CaI at above 50 GPa by employing first-principles structural searching and successfully identified the phase at 84 GPa using in situ X-ray diffraction. We further reveal that, due to the effect of orbital broadening, unusual charge transfer from the 5p orbitals of I to the 3d orbitals of Ca in CaI, gradually reverses the ionicity of Ca and becomes the anionic ICa at 485 GPa. Multivalent Ca stabilizes a set of metallic iodides with eight- to ten-fold iodine hyper-coordination. Our findings demonstrate that the valence states of Ca can vary from negative to +2, suggesting much greater complexity of Ca chemistry under ultrahigh pressures.

cond-mat.mtrl-sci↗

arXiv2025

Launching Insights: A Pilot Study on Leveraging Real-World Observational Data from the Mayo Clinic Platform to Advance Clinical Research

Backgrounds: Artificial intelligence (AI) is transforming healthcare, yet translating AI models from theoretical frameworks to real-world clinical applications remains challenging. The Mayo Clinic Platform (MCP) was established to address these challenges by providing a scalable ecosystem that integrates real-world multiple modalities data from multiple institutions, advanced analytical tools, and secure computing environments to support clinical research and AI development. Methods: In this study, we conducted four research projects leveraging MCP's data infrastructure and analytical capabilities to demonstrate its potential in facilitating real-world evidence generation and AI-driven clinical insights. Utilizing MCP's tools and environment, we facilitated efficient cohort identification, data extraction, and subsequent statistical or AI-powered analyses. Results: The results underscore MCP's role in accelerating translational research by offering de-identified, standardized real-world data and facilitating AI model validation across diverse healthcare settings. Compared to Mayo's internal Electronic Health Record (EHR) data, MCP provides broader accessibility, enhanced data standardization, and multi-institutional integration, making it a valuable resource for both internal and external researchers. Conclusion: Looking ahead, MCP is well-positioned to transform clinical research through its scalable ecosystem, effectively bridging the divide between AI innovation and clinical deployment. Future investigations will build upon this foundation, further exploring MCP's capacity to advance precision medicine and enhance patient outcomes.

cs.CY↗

arXiv2024

Prompting Large Language Models for Clinical Temporal Relation Extraction

Objective: This paper aims to prompt large language models (LLMs) for clinical temporal relation extraction (CTRE) in both few-shot and fully supervised settings. Materials and Methods: This study utilizes four LLMs: Encoder-based GatorTron-Base (345M)/Large (8.9B); Decoder-based LLaMA3-8B/MeLLaMA-13B. We developed full (FFT) and parameter-efficient (PEFT) fine-tuning strategies and evaluated these strategies on the 2012 i2b2 CTRE task. We explored four fine-tuning strategies for GatorTron-Base: (1) Standard Fine-Tuning, (2) Hard-Prompting with Unfrozen LLMs, (3) Soft-Prompting with Frozen LLMs, and (4) Low-Rank Adaptation (LoRA) with Frozen LLMs. For GatorTron-Large, we assessed two PEFT strategies-Soft-Prompting and LoRA with Frozen LLMs-leveraging Quantization techniques. Additionally, LLaMA3-8B and MeLLaMA-13B employed two PEFT strategies: LoRA strategy with Quantization (QLoRA) applied to Frozen LLMs using instruction tuning and standard fine-tuning. Results: Under fully supervised settings, Hard-Prompting with Unfrozen GatorTron-Base achieved the highest F1 score (89.54%), surpassing the SOTA model (85.70%) by 3.74%. Additionally, two variants of QLoRA adapted to GatorTron-Large and Standard Fine-Tuning of GatorTron-Base exceeded the SOTA model by 2.36%, 1.88%, and 0.25%, respectively. Decoder-based models with frozen parameters outperformed their Encoder-based counterparts in this setting; however, the trend reversed in few-shot scenarios. Discussions and Conclusions: This study presented new methods that significantly improved CTRE performance, benefiting downstream tasks reliant on CTRE systems. The findings underscore the importance of selecting appropriate models and fine-tuning strategies based on task requirements and data availability. Future work will explore larger models and broader CTRE applications.

cs.CL↗

arXiv2024

Improving Entity Recognition Using Ensembles of Deep Learning and Fine-tuned Large Language Models: A Case Study on Adverse Event Extraction from Multiple Sources

Adverse event (AE) extraction following COVID-19 vaccines from text data is crucial for monitoring and analyzing the safety profiles of immunizations. Traditional deep learning models are adept at learning intricate feature representations and dependencies in sequential data, but often require extensive labeled data. In contrast, large language models (LLMs) excel in understanding contextual information, but exhibit unstable performance on named entity recognition tasks, possibly due to their broad but unspecific training. This study aims to evaluate the effectiveness of LLMs and traditional deep learning models in AE extraction, and to assess the impact of ensembling these models on performance. In this study, we utilized reports and posts from the VAERS (n=621), Twitter (n=9,133), and Reddit (n=131) as our corpora. Our goal was to extract three types of entities: "vaccine", "shot", and "ae". We explored and fine-tuned (except GPT-4) multiple LLMs, including GPT-2, GPT-3.5, GPT-4, and Llama-2, as well as traditional deep learning models like RNN and BioBERT. To enhance performance, we created ensembles of the three models with the best performance. For evaluation, we used strict and relaxed F1 scores to evaluate the performance for each entity type, and micro-average F1 was used to assess the overall performance. The ensemble model achieved the highest performance in "vaccine", "shot", and "ae" with strict F1-scores of 0.878, 0.930, and 0.925, respectively, along with a micro-average score of 0.903. In conclusion, this study demonstrates the effectiveness and robustness of ensembling fine-tuned traditional deep learning models and LLMs, for extracting AE-related information. This study contributes to the advancement of biomedical natural language processing, providing valuable insights into improving AE extraction from text data for pharmacovigilance and public health surveillance.

cs.CL↗

arXiv2024

Relation Extraction Using Large Language Models: A Case Study on Acupuncture Point Locations

In acupuncture therapy, the accurate location of acupoints is essential for its effectiveness. The advanced language understanding capabilities of large language models (LLMs) like Generative Pre-trained Transformers (GPT) present a significant opportunity for extracting relations related to acupoint locations from textual knowledge sources. This study aims to compare the performance of GPT with traditional deep learning models (Long Short-Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers for Biomedical Text Mining (BioBERT)) in extracting acupoint-related location relations and assess the impact of pretraining and fine-tuning on GPT's performance. We utilized the World Health Organization Standard Acupuncture Point Locations in the Western Pacific Region (WHO Standard) as our corpus, which consists of descriptions of 361 acupoints. Five types of relations ('direction_of,' 'distance_of,' 'part_of,' 'near_acupoint,' and 'located_near') (n= 3,174) between acupoints were annotated. Five models were compared: BioBERT, LSTM, pre-trained GPT-3.5, fine-tuned GPT-3.5, as well as pre-trained GPT-4. Performance metrics included micro-average exact match precision, recall, and F1 scores. Our results demonstrate that fine-tuned GPT-3.5 consistently outperformed other models in F1 scores across all relation types. Overall, it achieved the highest micro-average F1 score of 0.92. This study underscores the effectiveness of LLMs like GPT in extracting relations related to acupoint locations, with implications for accurately modeling acupuncture knowledge and promoting standard implementation in acupuncture training and practice. The findings also contribute to advancing informatics applications in traditional and complementary medicine, showcasing the potential of LLMs in natural language processing.

cs.CL↗

arXiv2023

AE-GPT: Using Large Language Models to Extract Adverse Events from Surveillance Reports-A Use Case with Influenza Vaccine Adverse Events

Though Vaccines are instrumental in global health, mitigating infectious diseases and pandemic outbreaks, they can occasionally lead to adverse events (AEs). Recently, Large Language Models (LLMs) have shown promise in effectively identifying and cataloging AEs within clinical reports. Utilizing data from the Vaccine Adverse Event Reporting System (VAERS) from 1990 to 2016, this study particularly focuses on AEs to evaluate LLMs' capability for AE extraction. A variety of prevalent LLMs, including GPT-2, GPT-3 variants, GPT-4, and Llama 2, were evaluated using Influenza vaccine as a use case. The fine-tuned GPT 3.5 model (AE-GPT) stood out with a 0.704 averaged micro F1 score for strict match and 0.816 for relaxed match. The encouraging performance of the AE-GPT underscores LLMs' potential in processing medical data, indicating a significant stride towards advanced AE detection, thus presumably generalizable to other AE extraction tasks.

cs.CL↗

arXiv2020

Robustly Pre-trained Neural Model for Direct Temporal Relation Extraction

Background: Identifying relationships between clinical events and temporal expressions is a key challenge in meaningfully analyzing clinical text for use in advanced AI applications. While previous studies exist, the state-of-the-art performance has significant room for improvement. Methods: We studied several variants of BERT (Bidirectional Encoder Representations using Transformers) some involving clinical domain customization and the others involving improved architecture and/or training strategies. We evaluated these methods using a direct temporal relations dataset which is a semantically focused subset of the 2012 i2b2 temporal relations challenge dataset. Results: Our results show that RoBERTa, which employs better pre-training strategies including using 10x larger corpus, has improved overall F measure by 0.0864 absolute score (on the 1.00 scale) and thus reducing the error rate by 24% relative to the previous state-of-the-art performance achieved with an SVM (support vector machine) model. Conclusion: Modern contextual language modeling neural networks, pre-trained on a large corpus, achieve impressive performance even on highly-nuanced clinical temporal relation tasks.

cs.CL↗

arXiv2019

Reverse chemistry of iron in the deep Earth

In this work, we demonstrate a remarkable change of chemical trend of Iron under high pressure that is of great importance for understanding the distribution of elements in the Earth's mantle and core. Using first principles crystal structure search method, we conduct a systematic study of the propensity of p block elements to chemically bind with iron under high pressures ranging from ambient conditions to that of Earth's core. We show that under increasing pressure, iron tends to reverse its chemical nature, changing from an electron donor (reductant) to an electron acceptor, and oxidizes p-block elements in many compounds. Such reverse chemistry has a significant impact on the stoichiometries, bond types and strengths, structures and properties of iron compounds under deep planetary conditions.

physics.chem-ph↗

arXiv2018

Half-metallicity in honeycomb-kagome-lattice Mg3C2 monolayer with carrier doping

To obtain high-performance spintronic devices with high integration density, two-dimensional (2D) half-metallic materials are eagerly pursued all along. Here, we propose a stable 2D material with a honeycomb-kagome lattice, i.e., the Mg3C2 monolayer, based on first-principles calculations. This monolayer is an anti-ferromagnetic (AFM) semiconductor at its ground state. We further demonstrate that a transition from AFM semiconductor to ferromagnetic half-metal in this 2D material can be induced by carrier (electron or hole) doping. This magnetic transition can be understood by the Stoner criterion. In addition, the half-metallicity arises from the 2pz orbitals of the carbon (C) atoms for the electron-doped system, but from the C 2px and 2py orbitals for the case of hole doping. Our findings highlight a new promising material with controllable magnetic and electronic properties toward 2D spintronic applications.

cond-mat.mtrl-sci↗

arXiv2017

C$_4$N$_3$H monolayer: A novel two-dimensional organic Dirac material with high Fermi velocity

Searching for two-dimensional (2D) organic Dirac materials, which have more adaptable practical applications in comparing with inorganic ones, is of great significance and has been ongoing. However, only two kinds of these materials with low Fermi velocity have been discovered so far. Herein, we report the design of an organic monolayer with C$_4$N$_3$H stoichiometry which possesses fascinating structure and good stability in its free-standing state. More importantly, we demonstrate that this monolayer is a semimetal with anisotropic Dirac cones and very high Fermi velocity. This Fermi velocity is roughly one order of magnitude larger than that in 2D organic Dirac materials ever reported, and is comparable to that in graphene. The Dirac states in this monolayer arise from the extended $π$-electron conjugation system formed by the overlapping 2\emph{p}$_z$ orbitals of carbon and nitrogen atoms. Our finding opens a door for searching more 2D organic Dirac materials with high Fermi velocity.

cond-mat.mtrl-sci↗

arXiv2015

CH-π interaction-induced deep orbital deformation in a benzene-methane weak binding system

The nonbonding interaction between benzene and methane, called CH-π interaction, plays an important role in physical, chemical, and biological fields. CH-π interaction can decrease the system total energy and promote the formation of special geometric configurations. This work investigates systemically the orbital distribution and composition of the benzene-methane complex for the first time using ab initio calculation based on different methods and basis sets. Surprisingly, we find strong deformation in HOMO-4 and LUMO+2 induced by CH-π interaction, extending the general view that nonbonding interaction does not cause orbital change of molecules.

physics.chem-ph↗

arXiv2012

Diamondoid Structure of Polymeric Nitrogen at High Pressures

High-pressure polymeric structures of nitrogen have attracted great attention owing to their potential application as high-energy-density materials. We report the density functional structural prediction of the unexpected stabilization of a diamondoid (or N10-cage) structure of polymeric nitrogen at high pressures. The structure adopts a highly symmetric body-centered cubic form with lattice sites occupied by N10 tetracyclic cages, each of which consists of 10 atoms and is covalently bonded with its six next-nearest N10 cages. The prediction of this diamondoid structure rules out the earlier proposed helical tunnel phase and demonstrates the high-order nature of polymeric nitrogen at extreme high pressures. Diamondoid nitrogen is a wide-gap insulator and energetically more favorable than the experimental cubic gauche and previously predicted layered Pba2 phases above 263 GPa, a pressure which is accessible to high pressure experiment.

cond-mat.mtrl-sci↗