# DeepStack: Deeply Stacking Visual Tokens is Surprisingly Simple and Effective for LMMs

arXiv: https://arxiv.org/abs/2406.04334 · HTML: https://mlsanity.com/paper/2406.04334/

## Key facts

- Cited by 75 papers in the map (data through 14 Oct 2026).
- Placed in Visual token pruning and efficiency, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 62 of its 81 references that are in the map.
- First posted on arXiv 6 Jun 2024.

## Placed in

- Multimodal LLMs / vision-language models / [Visual token pruning and efficiency](https://mlsanity.com/mllm/visual-token-pruning-and-efficiency/)

## Builds on (most cited first, 30 of 62)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [End-to-End Object Detection with Transformers](https://mlsanity.com/paper/2005.12872/) (2020)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension](https://mlsanity.com/paper/1910.13461/) (2019)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://mlsanity.com/paper/2102.05918/) (2021)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs](https://mlsanity.com/paper/1606.00915/) (2016)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://mlsanity.com/paper/1909.11942/) (2019)
- [MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models](https://mlsanity.com/paper/2304.10592/) (2023)
- [Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations](https://mlsanity.com/paper/1602.07332/) (2016)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [BLOOM: A 176B-Parameter Open-Access Multilingual Language Model](https://mlsanity.com/paper/2211.05100/) (2022)
- [Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone](https://mlsanity.com/paper/2404.14219/) (2024)
- [Evaluating Object Hallucination in Large Vision-Language Models](https://mlsanity.com/paper/2305.10355/) (2023)
- [Towards VQA Models That Can Read](https://mlsanity.com/paper/1904.08920/) (2019)
- [ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning](https://mlsanity.com/paper/2203.10244/) (2022)

## Built on by (most cited first, 30 of 75)

- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Magma: A Foundation Model for Multimodal AI Agents](https://mlsanity.com/paper/2502.13130/) (2025)
- [SlowFast-LLaVA: A Strong Training-Free Baseline for Video Large Language Models](https://mlsanity.com/paper/2407.15841/) (2024)
- [MM1.5: Methods, Analysis & Insights from Multimodal LLM Fine-tuning](https://mlsanity.com/paper/2409.20566/) (2024)
- [UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation](https://mlsanity.com/paper/2505.14682/) (2025)
- [STEP3-VL-10B Technical Report](https://mlsanity.com/paper/2601.09668/) (2026)
- [[CLS] Token Tells Everything Needed for Training-free Efficient MLLMs](https://mlsanity.com/paper/2412.05819/) (2024)
- [Multi-Layer Visual Feature Fusion in Multimodal LLMs: Methods, Analysis, and Best Practices](https://mlsanity.com/paper/2503.06063/) (2025)
- [TokenCarve: Information-Preserving Visual Token Compression in Multimodal Large Language Models](https://mlsanity.com/paper/2503.10501/) (2025)
- [UniVid: The Open-Source Unified Video Model](https://mlsanity.com/paper/2509.24200/) (2025)
- [FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Merging](https://mlsanity.com/paper/2602.08024/) (2026)
- [REDUCIO! Generating 1K Video within 16 Seconds using Extremely Compressed Motion Latents](https://mlsanity.com/paper/2411.13552/) (2024)
- [VMAD: Visual-enhanced Multimodal Large Language Model for Zero-Shot Anomaly Detection](https://mlsanity.com/paper/2409.20146/) (2024)
- [Qwen-RobotNav Technical Report: A Scalable Navigation Model Designed for an Agentic Navigation System](https://mlsanity.com/paper/2606.18112/) (2026)
- [FocusUI: Efficient UI Grounding via Position-Preserving Visual Token Selection](https://mlsanity.com/paper/2601.03928/) (2026)
- [TS-LLaVA: Constructing Visual Tokens through Thumbnail-and-Sampling for Training-Free Video Large Language Models](https://mlsanity.com/paper/2411.11066/) (2024)
- [Textual Steering Vectors Can Improve Visual Understanding in Multimodal Large Language Models](https://mlsanity.com/paper/2505.14071/) (2025)
- [Visual Jev: Accurate and Efficient Decisions from Shared Visual Context](https://mlsanity.com/paper/2609.25845/) (2026)
- [INST-IT: Boosting Instance Understanding via Explicit Visual Prompt Instruction Tuning](https://mlsanity.com/paper/2412.03565/) (2024)
- [From Pixels to Words -- Towards Native One-Vision Models at Scale](https://mlsanity.com/paper/2605.28820/) (2026)
- [Enhancing Vision Foundation Models via Multimodal Continual Pre-Training](https://mlsanity.com/paper/2503.18931/) (2025)
- [The Unseen Bias: How Norm Discrepancy in Pre-Norm MLLMs Leads to Visual Information Loss](https://mlsanity.com/paper/2512.08374/) (2025)
- [VideoLoom: A Video Large Language Model for Joint Spatial-Temporal Understanding](https://mlsanity.com/paper/2601.07290/) (2026)
- [Looping Back to Move Forward: Recursive Transformers for Efficient and Flexible Large Multimodal Models](https://mlsanity.com/paper/2602.09080/) (2026)
- [U-VLM: Hierarchical Vision Language Modeling for Report Generation](https://mlsanity.com/paper/2603.00479/) (2026)
- [INF-LLaVA: Dual-perspective Perception for High-Resolution Multimodal Large Language Model](https://mlsanity.com/paper/2407.16198/) (2024)
- [Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs](https://mlsanity.com/paper/2605.00814/) (2026)
- [Earth-OneVision: Extending Remote Sensing Multimodal Large Language Models to More Sensor Modalities and Tasks](https://mlsanity.com/paper/2606.10819/) (2026)
- [Unified Multimodal Autoregressive Modeling with Shared Context-Visual Tokenizer is Key to Unification](https://mlsanity.com/paper/2606.18249/) (2026)
- [METEOR: Multi-Encoder Collaborative Token Pruning for Efficient Vision Language Models](https://mlsanity.com/paper/2507.20842/) (2025)

## Abstract (arXiv)

Most large multimodal models (LMMs) are implemented by feeding visual tokens as a sequence into the first layer of a large language model (LLM). The resulting architecture is simple but significantly increases computation and memory costs, as it has to handle a large number of additional tokens in its input layer. This paper presents a new architecture DeepStack for LMMs. Considering N layers in the language and vision transformer of LMMs, we stack the visual tokens into N groups and feed each group to its aligned transformer layer textitfrom bottom to top. Surprisingly, this simple method greatly enhances the power of LMMs to model interactions among visual tokens across layers but with minimal additional cost. We apply DeepStack to both language and vision transformer in LMMs, and validate the effectiveness of DeepStack LMMs with extensive empirical results. Using the same context length, our DeepStack 7B and 13B parameters surpass their counterparts by 2.7 and 2.9 on average across 9 benchmarks, respectively. Using only one-fifth of the context length, DeepStack rivals closely to the counterparts that use the full context length. These gains are particularly pronounced on high-resolution tasks, e.g., 4.2, 11.0, and 4.0 improvements on TextVQA, DocVQA, and InfoVQA compared to LLaVA-1.5-7B, respectively. We further apply DeepStack to vision transformer layers, which brings us a similar amount of improvements, 3.8 on average compared with LLaVA-1.5-7B.