ID-Align: RoPE-Conscious Position Remapping for Dynamic High-Resolution Adaptation in Vision-Language Models

arXiv:2505.21465 · cs.CV, cs.CL · Submitted 2025-05-27 · Read on arXiv

cs.CV, cs.CL

Submitted: 2025-05-27

Updated: 2026-09-08

Comments: Camera-ready version for EMNLP 2026 Findings

Code: https://github.com/zooblastlbz/ID-Align

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Embeddings — 15× in this paper · explained in 2 episodes
• Benchmark — 12× in this paper · explained in 2 episodes
• Transformer — 9× in this paper · explained in 2 episodes
• Vision-language models — 7× in this paper · explained in 2 episodes
• Adaptation — 4× in this paper · explained in 1 episode
• Hallucination — 3× in this paper · explained in 3 episodes
• Attention Mechanism — 2× in this paper · explained in 3 episodes
• Instruction Tuning — 2× in this paper · explained in 2 episodes
• Vision-Language Models (VLMs) — 2× in this paper · explained in 6 episodes
• Adam — 1× in this paper · explained in 2 episodes
• Benchmarking — 1× in this paper · explained in 3 episodes
• LLM-as-a-Judge — 1× in this paper · explained in 3 episodes
• Multimodal Reasoning — 1× in this paper · explained in 5 episodes
• Test-time Scaling — 1× in this paper · explained in 2 episodes
• Transfer Learning — 1× in this paper · explained in 10 episodes

Sources

• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
• GPT-4 Technical Report
• NVLM: Open Frontier-Class Multimodal LLMs
• A Survey of Multimodal Large Language Model from A Data-centric Perspective
• Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context
• Round and Round We Go! What makes Rotary Positional Encodings useful?
• Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
• InternLM2 Technical Report
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• V2PE: Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding
• Are We on the Right Way for Evaluating Large Vision-Language Models?
• DeBERTa: Decoding-enhanced BERT with Disentangled Attention
• Advancing General Multimodal Capability of Vision-language Models with Pyramid-descent Visual Position Encoding
• SEED-Bench-2-Plus: Benchmarking Multimodal Large Language Models with Text-Rich Visual Comprehension
• Base of RoPE Bounds Context Length
• Pointer Sentinel Mixture Models
• Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation
• DeepSeek-V3 Technical Report
• Self-Attention with Relative Position Representations
• Encoding word order in complex embeddings

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models