Jie Wang
YOU?
Author Swipe
View article: Data Whisperer: Efficient Data Selection for Task-Specific LLM Fine-Tuning via Few-Shot In-Context Learning
Data Whisperer: Efficient Data Selection for Task-Specific LLM Fine-Tuning via Few-Shot In-Context Learning Open
Fine-tuning large language models (LLMs) on task-specific data is essential for their effective deployment. As dataset sizes grow, efficiently selecting optimal subsets for training becomes crucial to balancing performance and computationa…
View article: Case Report: A novel chemotherapy-free regimen combined with photodynamic therapy, target therapy, and immunotherapy in a geriatric male with huge recurrent scalp and facial angiosarcoma: a report of an extremely rare case and literature review
Case Report: A novel chemotherapy-free regimen combined with photodynamic therapy, target therapy, and immunotherapy in a geriatric male with huge recurrent scalp and facial angiosarcoma: a report of an extremely rare case and literature review Open
Angiosarcomas are sporadic vascular neoplasms among the most aggressive subtypes of soft tissue sarcomas. In addition, vast and multiple recurrent superficial scalp and facial angiosarcomas are very complex and extremely difficult to manag…
View article: SAIL: Sample-Centric In-Context Learning for Document Information Extraction
SAIL: Sample-Centric In-Context Learning for Document Information Extraction Open
Document Information Extraction (DIE) aims to extract structured information from Visually Rich Documents (VRDs). Previous full-training approaches have demonstrated strong performance but may struggle with generalization to unseen data. I…
View article: AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Society
AgentSociety: Large-Scale Simulation of LLM-Driven Generative Agents Advances Understanding of Human Behaviors and Society Open
Understanding human behavior and society is a central focus in social sciences, with the rise of generative social science marking a significant paradigmatic shift. By leveraging bottom-up simulations, it replaces costly and logistically c…
View article: Towards regional-scale overgrazing risk mapping in natural grasslands: a novel spectral–temporal fusion approach using Landsat and Sentinel-2 time series
Towards regional-scale overgrazing risk mapping in natural grasslands: a novel spectral–temporal fusion approach using Landsat and Sentinel-2 time series Open
View article: SAIL: Sample-Centric In-Context Learning for Document Information Extraction
SAIL: Sample-Centric In-Context Learning for Document Information Extraction Open
Document Information Extraction (DIE) aims to extract structured information from Visually Rich Documents (VRDs). Previous full-training approaches have demonstrated strong performance but may struggle with generalization to unseen data. I…
View article: Foulant Characteristics Comparison in Recycling Cooling Water System Makeup by Municipal Reclaimed Water and Surface Water in Power Plant
Foulant Characteristics Comparison in Recycling Cooling Water System Makeup by Municipal Reclaimed Water and Surface Water in Power Plant Open
Due to water shortage, municipal reclaimed water rather than surface water was replenished into recycling cooling water system in power plants in some cities in China. In order to understand the effects of the measure on carbon steel corro…