J. Q. Zhang
YOU?
Author Swipe
View article: Learning Dynamics of VLM Finetuning
Learning Dynamics of VLM Finetuning Open
Preference-based finetuning of vision--language models (VLMs) is brittle: trivially wrong negatives inject uninformative gradients that destabilize training. We recast alignment as \textbf{learning-dynamics--aware optimization} and introdu…
View article: GAM-Agent: Game-Theoretic and Uncertainty-Aware Collaboration for Complex Visual Reasoning
GAM-Agent: Game-Theoretic and Uncertainty-Aware Collaboration for Complex Visual Reasoning Open
We propose GAM-Agent, a game-theoretic multi-agent framework for enhancing vision-language reasoning. Unlike prior single-agent or monolithic models, GAM-Agent formulates the reasoning process as a non-zero-sum game between base agents--ea…
View article: A cartographic generalization method for 3D visualization of trajectories in space–time cubes: case study of epidemic spread
A cartographic generalization method for 3D visualization of trajectories in space–time cubes: case study of epidemic spread Open
The widespread adoption of positioning technology and location-based services has resulted in the continuous generation of substantial volumes of accessible spatiotemporal trajectory data. While many studies focus on 2D trajectory visualiz…
View article: A Study on Blueberry Variety Classification Based on the HMS-ResNeXt50 Model
A Study on Blueberry Variety Classification Based on the HMS-ResNeXt50 Model Open