# SA-GEM: Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning for Efficient Remote Sensing Large Vision-Language Models

arXiv: https://arxiv.org/abs/2608.15075 · HTML: https://mlsanity.com/paper/2608.15075/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Embodied and 3D vision-language models, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 20 of its 26 references that are in the map.
- First posted on arXiv 15 Aug 2026.

## Placed in

- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- Efficient LLM inference and serving / [Visual token pruning in multimodal models](https://mlsanity.com/inference/visual-token-pruning-in-multimodal-models/)

## Builds on (most cited first, 20 of 20)

- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [Token Merging: Your ViT But Faster](https://mlsanity.com/paper/2210.09461/) (2022)
- [GeoChat: Grounded Large Vision-Language Model for Remote Sensing](https://mlsanity.com/paper/2311.15826/) (2023)
- [PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction](https://mlsanity.com/paper/2410.17247/) (2024)
- [MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?](https://mlsanity.com/paper/2408.13257/) (2024)
- [Stop Looking for Important Tokens in Multimodal Language Models: Duplication Matters More](https://mlsanity.com/paper/2502.11494/) (2025)
- [ClearCLIP: Decomposing CLIP Representations for Dense Vision-Language Inference](https://mlsanity.com/paper/2407.12442/) (2024)
- [Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areas](https://mlsanity.com/paper/2503.01773/) (2025)
- [VHM: Versatile and Honest Vision Language Model for Remote Sensing Image Analysis](https://mlsanity.com/paper/2403.20213/) (2024)
- [EarthDial: Turning Multi-sensory Earth Observations to Interactive Dialogues](https://mlsanity.com/paper/2412.15190/) (2024)
- [GeoLLaVA-8K: Scaling Remote-Sensing Multimodal Large Language Models to 8K Resolution](https://mlsanity.com/paper/2505.21375/) (2025)
- [When Large Vision-Language Model Meets Large Remote Sensing Imagery: Coarse-to-Fine Text-Guided Token Pruning](https://mlsanity.com/paper/2503.07588/) (2025)
- [FOLDER: Accelerating Multi-modal Large Language Models with Enhanced Performance](https://mlsanity.com/paper/2501.02430/) (2025)
- [RSUniVLM: A Unified Vision Language Model for Remote Sensing via Granularity-oriented Mixture of Experts](https://mlsanity.com/paper/2412.05679/) (2024)
- [ZoomEarth: Active Perception for Ultra-High-Resolution Geospatial Vision-Language Tasks](https://mlsanity.com/paper/2511.12267/) (2025)
- [A Benchmark for Ultra-High-Resolution Remote Sensing MLLMs](https://mlsanity.com/paper/2512.17319/) (2025)
- [SkyMoE: A Vision-Language Foundation Model for Enhancing Geospatial Interpretation with Mixture of Experts](https://mlsanity.com/paper/2512.02517/) (2025)
- [Improving Visual Token Reduction via Rectifying Distortions for Efficient Multimodal LLM Inference](https://mlsanity.com/paper/2606.01711/) (2026)

## Abstract (arXiv)

RS-LVLMs have advanced multimodal understanding of Earth observation imagery, yet their performance is fundamentally constrained by high-resolution processing, as visual token counts grow quadratically with linear input resolution while important visual evidence is inherently sparse and increasingly diluted across the expanded sequence. Existing token pruning methods largely rely on scale-agnostic resolution policies and isolated importance cues, limiting task-aligned granularity adaptation and holistic evidence preservation. To address this, we present Scale-Adaptive and Geospatial Evidence-Modulated Token Pruning (SA-GEM), a plug-and-play framework that unifies task-adaptive token granularity allocation with holistic geospatial token importance modulation. Specifically, a lightweight router selects the resolution based on query-dependent token granularity, while a token importance modulator jointly models task relevance, spatial structure, and local redundancy to preserve holistic geospatial evidence. We show that higher resolution is not universally beneficial and, once sufficient granularity is reached, token quality matters more than token quantity. Experiments across various benchmarks demonstrate that SA-GEM achieves consistent gains in both accuracy and efficiency over existing pruning methods. On XLRS-Bench, it surpasses GeoLLaVA-8K by 2.3% in accuracy with a 2.4 times total inference speedup.