Qing Li

Introducing a new method to reduce memory use when finetuning large models.

A new dataset enhances the connection between language and 3D environments.

A new method enhances molecule-caption translation using Large Language Models.

A novel approach for better interpreting videos and answering complex questions.

A look at using active learning for enhanced generative model performance.

BDoG improves AI reasoning by integrating various data types effectively.

This paper explores skill acquisition patterns using machine learning techniques on large datasets.

Discover the potential of personalized image generation techniques.

Research reveals new insights into superconductivity and charge density waves in Cs(VTa)Sb.

Research reveals exciting features in Ta-CsVSb kagome superconductors.

LLM4EA enhances the efficiency of connecting entities in diverse knowledge graphs.

An overview of how language models like Transformers operate and their significance.

A new framework enhances the understanding of hidden sentiments in text.

Discover the latest in computer vision powered by deep learning models.

A new approach combines LLMs and pseudo-events for better video highlights.

Introducing PolyRAG, a system that enhances accuracy in information retrieval across specialized fields.

New methods aim to enhance the speed and efficiency of deep learning models.

Mamba offers a new architecture for efficient handling of complex data in AI.

Study reveals effective methods to identify hallucinations in large vision-language models.

SSD4Rec offers a fresh approach to personalized recommendations for users.

Exploring how memory functions in LLMs and its comparison to human memory.

A model to assess segmentation quality without ground truth benchmarks.

Discover how models can learn from errors in visual reasoning.

A look into the advancements of implicit sentiment analysis using innovative frameworks.

TOMG-Bench revolutionizes how language models assist scientists in creating new molecules.