mlsanityData to 7 Oct 2026

Revisiting Multimodal Positional Encoding in Vision-Language Models

arXiv 2510.23095 ↗First posted 27 Oct 2025cs.CVcited by 13

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

Abstract (arXiv)

Multimodal position encoding is essential for vision-language models, yet there has been little systematic investigation into multimodal position encoding. We conduct a comprehensive analysis of multimodal Rotary Positional Embedding (RoPE) by examining its two core components: position design and frequency allocation. Through extensive experiments, we identify three key guidelines: positional coherence, full frequency utilization, and preservation of textual priors-ensuring unambiguous layout, rich representation, and faithful transfer from the pre-trained LLM. Based on these insights, we propose Multi-Head RoPE (MHRoPE) and MRoPE-Interleave (MRoPE-I), two simple and plug-and-play variants that require no architectural changes. Our methods consistently outperform existing approaches across diverse benchmarks, with significant improvements in both general and fine-grained multimodal understanding. Code will be avaliable at https://github.com/JJJYmmm/Multimodal-RoPEs.

Key facts. Cited by 13 papers in the map (data through 7 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 27 of its 34 references that are in the map. First posted on arXiv 27 Oct 2025.

Builds on

27 of 34 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Circle-RoPE: Cone-like Decoupled Rotary Positional Embedding for Large Vision-Language Models22 May
  2. 2024V2PE: Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding12 Dec
  3. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution18 Sep
  4. 2024BLINK: Multimodal Large Language Models Can See but Not Perceive18 Apr
  5. 2023MMBench: Is Your Multi-modal Model an All-around Player?12 Jul

Other references · 22

  1. 2017Attention Is All You Needcited by 59,932
  2. 2025Qwen3 Technical Reportcited by 8,195
  3. 2025Qwen2.5-VL Technical Reportcited by 5,418
  4. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  5. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
  6. 2019Towards VQA Models That Can Readcited by 1,696
  7. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoningcited by 1,449
  8. 2024Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysiscited by 1,375
  9. 2020DocVQA: A Dataset for VQA on Document Imagescited by 1,288
  10. 2023MVBench: A Comprehensive Multi-modal Video Understanding Benchmarkcited by 1,086
  11. 2024Are We on the Right Way for Evaluating Large Vision-Language Models?cited by 893
  12. 2016A Diagram Is Worth A Dozen Imagescited by 793
10 more
  1. 2023YaRN: Efficient Context Window Extension of Large Language Modelscited by 584
  2. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 466
  3. 2021InfographicVQAcited by 430
  4. 2024LVBench: An Extreme Long Video Understanding Benchmarkcited by 402
  5. 2025OmniGen2: Towards Instruction-Aligned Multimodal Generationcited by 241
  6. 2024STAR: A Benchmark for Situated Reasoning in Real-World Videoscited by 213
  7. 2025Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generationcited by 108
  8. 2024Apollo: An Exploration of Video Understanding in Large Multimodal Modelscited by 87
  9. 2025VideoRoPE: What Makes for Good Video Rotary Position Embedding?cited by 63
  10. 2025VRoPE: Rotary Position Embedding for Video Large Language Modelscited by 17

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 12.

13 papers.

Sort:

By area:

  1. 2026DVGT-2: Vision-Geometry-Action Model for Autonomous Driving at Scalecited by 12
  2. 2026FocusUI: Efficient UI Grounding via Position-Preserving Visual Token Selectioncited by 9
  3. 2026Think-as-You-See: Streaming Chain-of-Thought Reasoning for Large Vision-Language Modelscited by 6
  4. 2025EchoingPixels: Aliasing-Resistant Joint Token Reduction for Audio-Visual LLMscited by 4
  5. 2026Beyond Surface Artifacts: Capturing Shared Latent Forgery Knowledge Across Modalitiescited by 4
  6. 2026Linking Perception, Confidence and Accuracy in MLLMscited by 2
  7. 2026Qwen-3D: A Generalist 3D Vision-Language Model for Spatial Understandingcited by 2
  8. 2026Uncovering Understanding-Generation Synergy in Native Unified Multimodal Models: From Representation, Task to Systemcited by 2
  9. 2026Beyond Sequential Distance: Inter-Modal Distance Invariant Position Encodingcited by 1
  10. 2026Unveiling the Visual Counting Bottleneck in Vision-Language Modelscited by 1
  11. 2026InCaRPose: In-Cabin Relative Camera Pose Estimation Model and Datasetcited by 0
  12. 2026RIG-RoPE: Relation-Stratified Multimodal Attention with Instance-Local Rotary Geometry and Representation-Aware Traversal Coordinatescited by 0
1 more
  1. 2026Beyond Compression: Training Latent Representations for Stable Long-Horizon Rollout in Neural Surrogate Solverscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025HoPE: Hybrid of Position Embedding for Long Context Vision-Language Models
  2. 2025OMEGA: Optimized Multimodal Position Encoding Index Derivation with Global Adaptive Scaling for Vision-Language Models
  3. 2025Advancing General Multimodal Capability of Vision-language Models with Pyramid-descent Visual Position Encoding
  4. 2026C^2ROPE: Causal Continuous Rotary Positional Encoding for 3D Large Multimodal-Models Reasoning
  5. 2025Unified Multimodal Understanding via Byte-Pair Visual Encoding
  6. 2026MODIX: A Training-Free Multimodal Information-Driven Positional Index Scaling for Vision-Language Models
  7. 2026Layer-Aware Position Embeddings for Visual Token Pruning in Multimodal Large Language Models
  8. 2025Beyond Semantics: Rediscovering Spatial Awareness in Vision-Language Models