# M2IV: Towards Efficient and Fine-grained Multimodal In-Context Learning via Representation Engineering

arXiv: https://arxiv.org/abs/2504.04633 · HTML: https://mlsanity.com/paper/2504.04633/

## Key facts

- Cited by 35 papers in the map (data through 14 Oct 2026).
- Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 44 of its 55 references that are in the map.
- First posted on arXiv 6 Apr 2025.

## Placed in

- Multimodal LLMs / vision-language models / [MLLM training and visual instruction tuning](https://mlsanity.com/mllm/mllm-training-and-visual-instruction-tuning/)
- Mechanistic interpretability / [Interpreting vision and multimodal models](https://mlsanity.com/mechinterp/interpreting-vision-and-multimodal-models/)

## Builds on (most cited first, 30 of 44)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Learning to Prompt for Vision-Language Models](https://mlsanity.com/paper/2109.01134/) (2021)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge](https://mlsanity.com/paper/1906.00067/) (2019)
- [An Explanation of In-context Learning as Implicit Bayesian Inference](https://mlsanity.com/paper/2111.02080/) (2021)
- [VizWiz Grand Challenge: Answering Visual Questions from Blind People](https://mlsanity.com/paper/1802.08218/) (2018)
- [Vision-Language Models for Vision Tasks: A Survey](https://mlsanity.com/paper/2304.00685/) (2023)
- [What Can Transformers Learn In-Context? A Case Study of Simple Function Classes](https://mlsanity.com/paper/2208.01066/) (2022)
- [A-OKVQA: A Benchmark for Visual Question Answering using World Knowledge](https://mlsanity.com/paper/2206.01718/) (2022)
- [OpenFlamingo: An Open-Source Framework for Training Large Autoregressive Vision-Language Models](https://mlsanity.com/paper/2308.01390/) (2023)
- [The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes](https://mlsanity.com/paper/2005.04790/) (2020)
- [What matters when building vision-language models?](https://mlsanity.com/paper/2405.02246/) (2024)
- [OBELICS: An Open Web-Scale Filtered Dataset of Interleaved Image-Text Documents](https://mlsanity.com/paper/2306.16527/) (2023)
- [In-Context Learning Creates Task Vectors](https://mlsanity.com/paper/2310.15916/) (2023)
- [Function Vectors in Large Language Models](https://mlsanity.com/paper/2310.15213/) (2023)
- [Long-CLIP: Unlocking the Long-Text Capability of CLIP](https://mlsanity.com/paper/2403.15378/) (2024)
- [MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models](https://mlsanity.com/paper/2311.17600/) (2023)
- [In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering](https://mlsanity.com/paper/2311.06668/) (2023)
- [Label Words are Anchors: An Information Flow Perspective for Understanding In-Context Learning](https://mlsanity.com/paper/2305.14160/) (2023)
- [Why Can GPT Learn In-Context? Language Models Implicitly Perform Gradient Descent as Meta-Optimizers](https://mlsanity.com/paper/2212.10559/) (2022)
- [Vision-Language Models for Medical Report Generation and Visual Question Answering: A Review](https://mlsanity.com/paper/2403.02469/) (2024)

## Built on by (most cited first, 30 of 35)

- [UniCTokens: Boosting Personalized Understanding and Generation via Unified Concept Tokens](https://mlsanity.com/paper/2505.14671/) (2025)
- [UniME-V2: MLLM-as-a-Judge for Universal Multimodal Embedding Learning](https://mlsanity.com/paper/2510.13515/) (2025)
- [TACO: Enhancing Multimodal In-context Learning via Task Mapping-Guided Sequence Configuration](https://mlsanity.com/paper/2505.17098/) (2025)
- [Causal Tracing of Object Representations in Large Vision Language Models: Mechanistic Interpretability and Hallucination Mitigation](https://mlsanity.com/paper/2511.05923/) (2025)
- [TEMA: Anchor the Image, Follow the Text for Multi-Modification Composed Image Retrieval](https://mlsanity.com/paper/2604.21806/) (2026)
- [Steering Vector Fields for Context-Aware Inference-Time Control in Large Language Models](https://mlsanity.com/paper/2602.01654/) (2026)
- [CentaurEval: Benchmarking Human-in-the-Loop Value in Agentic Coding](https://mlsanity.com/paper/2512.04111/) (2025)
- [Hy-Facial: Hybrid Feature Extraction by Dimensionality Reduction Methods for Enhanced Facial Expression Classification](https://mlsanity.com/paper/2509.26614/) (2025)
- [MRFD: Multi-Region Fusion Decoding with Self-Consistency for Mitigating Hallucinations in LVLMs](https://mlsanity.com/paper/2508.10264/) (2025)
- [Train Once, Reuse Everywhere: Generalizable Implicit In-Context Learning by Routing Attention](https://mlsanity.com/paper/2509.22854/) (2025)
- [FakeSV-VLM: Taming VLM for Detecting Fake Short-Video News via Progressive Mixture-Of-Experts Adapter](https://mlsanity.com/paper/2508.19639/) (2025)
- [Instruction Anchor: Dissecting the Mechanistic Dynamics of Modality Arbitration](https://mlsanity.com/paper/2602.03677/) (2026)
- [ActorMind: Emulating Human Actor Reasoning for Speech Role-Playing](https://mlsanity.com/paper/2604.11103/) (2026)
- [U2UData+: A Scalable Swarm UAVs Autonomous Flight Dataset for Embodied Long-horizon Tasks](https://mlsanity.com/paper/2509.00055/) (2025)
- [HotComment: A Benchmark for Evaluating Popularity of Online Comments](https://mlsanity.com/paper/2604.25614/) (2026)
- [MM-StanceDet: Retrieval-Augmented Multi-modal Multi-agent Stance Detection](https://mlsanity.com/paper/2604.27934/) (2026)
- [Personalize Your Large Vision-language Models With In-context Prompt Tuning](https://mlsanity.com/paper/2605.31513/) (2026)
- [Enhancing Multimodal In-Context Learning via Inductive-Deductive Reasoning](https://mlsanity.com/paper/2605.02378/) (2026)
- [A2TGPO: Agentic Turn-Group Policy Optimization with Adaptive Turn-level Clipping](https://mlsanity.com/paper/2605.06200/) (2026)
- [STARE at the Structure: Steering ICL Exemplar Selection with Structural Alignment](https://mlsanity.com/paper/2508.20944/) (2025)
- [HIFICL: High-Fidelity In-Context Learning for Multimodal Tasks](https://mlsanity.com/paper/2603.12760/) (2026)
- [Balancing Multimodal Domain Generalization via Gradient Modulation and Projection](https://mlsanity.com/paper/2603.14175/) (2026)
- [Hyper-ICL: Attention Calibration with Hyperbolic Anchor Distillation for Multimodal In-Context Learning](https://mlsanity.com/paper/2606.04434/) (2026)
- [Beyond Surface Imitation: Contrastive Modeling for Reasoning Path Alignment in Multimodal In-Context Learning](https://mlsanity.com/paper/2609.10177/) (2026)
- [ModRWKV: Transformer Multimodality in Linear Time](https://mlsanity.com/paper/2505.14505/) (2025)
- [Zero-shot Cross-lingual NER via Mitigating Language Difference: An Entity-aligned Translation Perspective](https://mlsanity.com/paper/2509.01147/) (2025)
- [Two Heads are Better than One: Distilling Large Language Model Features Into Small Models with Feature Decomposition and Mixture](https://mlsanity.com/paper/2511.07110/) (2025)
- [UniCreative: Unifying Long-form Logic and Short-form Sparkle via Reference-Free Reinforcement Learning](https://mlsanity.com/paper/2604.05517/) (2026)
- [Neutral-Reference Prompting for Vision-Language Models](https://mlsanity.com/paper/2605.15615/) (2026)
- [Criterion-Conditional In-Context Learning: Evaluating Criterion-Shift Adaptation in Vision-Language Models](https://mlsanity.com/paper/2607.02575/) (2026)

## Abstract (arXiv)

Multimodal in-context learning (ICL) equips Large Vision-language Models (LVLMs) with the ability to adapt to new tasks via multiple user-provided demonstrations, without requiring any model parameter updates. However, its effectiveness is constrained by the token-intensive nature of multimodal inputs and the complexity of cross-modal few-shot reasoning, which together hinder LVLMs from extracting useful patterns from demonstrations. To address these challenges, we propose M2IV, a novel representation engineering approach that replaces explicit token-level demonstrations with a set of learnable Multimodal In-context Vectors directly injected into the residual streams of LVLMs. By analyzing the distinct roles of multi-head attention (MHA) and multi-layer perceptrons (MLP) in the ICL process, we design a training strategy that enables M2IV to perform fine-grained semantic distillation and robust cross-modal representation learning. M2IV not only improves performance across diverse tasks and LVLMs but also significantly reduces token overhead, enabling graceful scaling to many-shot scenarios. To further enhance usability, we introduce VLibrary, a repository that stores trained M2IVs for flexible retrieval and injection. With VLibrary, users can steer pre-trained LVLMs in a customized manner that meets diverse requirements. Extensive experiments demonstrate that M2IV consistently outperforms vanilla ICL and prior representation engineering baselines, achieving an average accuracy gain of 3.74% with substantial improvements in overall efficiency.