SkyNative: A Native Multimodal Framework for Remote Sensing Visual Evidence Reasoning
For remote sensing vision-language models, this work addresses the problem of over-reliance on language priors in fine-grained spatial reasoning, offering a more reliable approach.
SkyNative introduces an encoder-free multimodal framework for remote sensing that directly represents images as raw patch tokens, avoiding premature compression of visual evidence. It achieves stronger image-grounded perception and improved robustness against language priors in fine-grained spatial reasoning tasks.
Remote sensing vision-language models commonly rely on pretrained visual encoders to convert images into semantic features before language-model reasoning. While effective for scene-level understanding, this pipeline may prematurely compress local visual evidence, making fine-grained spatial reasoning vulnerable to language priors, especially in ultra-high-resolution remote sensing imagery. We present SkyNative, a native multimodal framework for remote sensing that adopts an encoder-free architecture, removing the pretrained visual backbone to directly represent images as raw patch tokens in the language-model token space. To reconcile low-level visual patches with textual tokens, SkyNative introduces a modality-aware decoupling mechanism that uses modality-specific parameters within a unified autoregressive backbone. We further introduce a visual reliance benchmark that diagnoses whether models ground their answers in image evidence through progressive visual degradation and misleading textual prompts. Across standard remote sensing understanding tasks and large-format spatial reasoning evaluations, SkyNative shows stronger image-grounded perception and improved robustness against prompt-induced language priors. These results suggest that native patch-level multimodal modeling is a promising direction for reliable remote sensing vision-language reasoning.