# Query Independent Variable Rate Visual Token Coding

arXiv: https://arxiv.org/abs/2610.00204 · HTML: https://mlsanity.com/paper/2610.00204/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Visual token pruning in multimodal models, one of 12 directions in Efficient LLM inference and serving.
- Builds on 11 of its 14 references that are in the map.
- First posted on arXiv 20 Sep 2026.

## Placed in

- 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, 11 of 11)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning](https://mlsanity.com/paper/2203.10244/) (2022)
- [Token Merging: Your ViT But Faster](https://mlsanity.com/paper/2210.09461/) (2022)
- [An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Models](https://mlsanity.com/paper/2403.06764/) (2024)
- [LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Models](https://mlsanity.com/paper/2403.15388/) (2024)
- [Hybrid Token Compression for Vision-Language Models](https://mlsanity.com/paper/2512.08240/) (2025)
- [Fourier Compressor: Frequency-Domain Visual Token Compression for Vision-Language Models](https://mlsanity.com/paper/2508.06038/) (2025)
- [OccamToken: Efficient VLM Inference with Training-Free and Budget-Adaptive Token Pruning](https://mlsanity.com/paper/2605.29657/) (2026)
- [Adaptive-VoCo: Complexity-Aware Visual Token Compression for Vision-Language Models](https://mlsanity.com/paper/2512.18496/) (2025)
- [RUTA: Principled Visual Token Allocation via Rate-Utility Optimization](https://mlsanity.com/paper/2608.04132/) (2026)

## Abstract (arXiv)

Visual-token compression for vision--language models is posed almost entirely as a selection problem: decide which tokens to keep and discard the rest. The criteria that work best rank tokens by the attention the language model pays them, which makes the ranking a function of the question being asked. That is invisible in a single-turn benchmark and decisive whenever a compressed representation is written once and read many times, as when it is cached across the turns of a conversation or transmitted between a device and a server. We take the other half of the classical transform-coding toolkit instead: keep every token and vary its rate. A transform code exposes each token's measured distortion--rate curve, and a fixed bit budget is distributed across tokens by exact integer rate--distortion optimisation on those curves. No text enters the pipeline, so one compressed representation serves any query. At equal bit budgets, on two datasets and two capacities, it preserves the model's output distribution and its answers better than uniform-rate coding, the closed-form water-fill and distortion-ranked pruning. It matches attention-ranked pruning on the question pruning was tuned for, and overtakes it once the compressed image must answer a different question about the same image.