4 papers
MODIX: A Training-Free Multimodal Information-Driven Positional Index Scaling for Vision-Language Models
Ruoxiang Huang, Zhen Yuan
Vision-Language Models (VLMs) have achieved remarkable progress in multimodal understanding, yet their positional encoding mechanisms remain suboptimal. Existing approaches uniform…
ID-LoRA: Efficient Low-Rank Adaptation Inspired by Matrix Interpolative Decomposition
Xindian Ma, Rundong Kong, Peng Zhang +2
LoRA has become a universal Parameter-Efficient Fine-Tuning (PEFT) technique that equips Large Language Models (LLMs) to adapt quickly to new tasks. However, when these models are…
OMEGA: Optimized Multimodal Position Encoding Index Derivation with Global Adaptive Scaling for Vision-Language Models
Ruoxiang Huang, Xindian Ma, Rundong Kong +2
Vision-Language Models (VLMs) have demonstrated strong performance across various multimodal tasks, where position encoding plays a vital role in modeling both the sequential struc…
DoPI: Doctor-like Proactive Interrogation LLM for Traditional Chinese Medicine
Zewen Sun, Ruoxiang Huang, Jiahe Feng +7
Enhancing interrogation capabilities in Traditional Chinese Medicine (TCM) diagnosis through multi-turn dialogues and knowledge graphs presents a significant challenge for modern A…