# VisNec: Measuring and Leveraging Visual Necessity for Multimodal Instruction Tuning

arXiv: https://arxiv.org/abs/2603.01195 · HTML: https://mlsanity.com/paper/2603.01195/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 37 of its 46 references that are in the map.
- First posted on arXiv 1 Mar 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)

## Builds on (most cited first, 30 of 37)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023)
- [MMBench: Is Your Multi-modal Model an All-around Player?](https://mlsanity.com/paper/2307.06281/) (2023)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [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)
- [MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models](https://mlsanity.com/paper/2306.13394/) (2023)
- [MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities](https://mlsanity.com/paper/2308.02490/) (2023)
- [Deep Learning on a Data Diet: Finding Important Examples Early in Training](https://mlsanity.com/paper/2107.07075/) (2021)
- [From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning](https://mlsanity.com/paper/2308.12032/) (2023)
- [Understanding Dataset Difficulty with V-Usable Information](https://mlsanity.com/paper/2110.08420/) (2021)
- [Active Learning on a Budget: Opposite Strategies Suit High and Low Budgets](https://mlsanity.com/paper/2202.02794/) (2022)
- [Vision-Flan: Scaling Human-Labeled Tasks in Visual Instruction Tuning](https://mlsanity.com/paper/2402.11690/) (2024)
- [Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection](https://mlsanity.com/paper/2402.12501/) (2024)
- [Multimodal Instruction Tuning with Conditional Mixture of LoRA](https://mlsanity.com/paper/2402.15896/) (2024)
- [MQuant: Unleashing the Inference Potential of Multimodal Large Language Models via Full Static Quantization](https://mlsanity.com/paper/2502.00425/) (2025)
- [Multimodal Spatial Reasoning in the Large Model Era: A Survey and Benchmarks](https://mlsanity.com/paper/2510.25760/) (2025)
- [ObjectRelator: Enabling Cross-View Object Relation Understanding Across Ego-Centric and Exo-Centric Perspectives](https://mlsanity.com/paper/2411.19083/) (2024)
- [ICONS: Influence Consensus for Vision-Language Data Selection](https://mlsanity.com/paper/2501.00654/) (2024)
- [Concept-skill Transferability-based Data Selection for Large Vision-Language Models](https://mlsanity.com/paper/2406.10995/) (2024)
- [XTrack: Multimodal Training Boosts RGB-X Video Object Trackers](https://mlsanity.com/paper/2405.17773/) (2024)
- [EgoCross: Benchmarking Multimodal Large Language Models for Cross-Domain Egocentric Video Question Answering](https://mlsanity.com/paper/2508.10729/) (2025)
- [EgoNight: Towards Egocentric Vision Understanding at Night with a Challenging Benchmark](https://mlsanity.com/paper/2510.06218/) (2025)
- [EarthMind: Leveraging Cross-Sensor Data for Advanced Earth Observation Interpretation with a Unified Multimodal LLM](https://mlsanity.com/paper/2506.01667/) (2025)
- [Vision-G1: Towards General Vision Language Reasoning with Multi-Domain Data Curation](https://mlsanity.com/paper/2508.12680/) (2025)
- [Filter Images First, Generate Instructions Later: Pre-Instruction Data Selection for Visual Instruction Tuning](https://mlsanity.com/paper/2503.07591/) (2025)
- [Yo'City: Personalized and Boundless 3D Realistic City Scene Generation via Self-Critic Expansion](https://mlsanity.com/paper/2511.18734/) (2025)
- [MICA: Multi-Agent Industrial Coordination Assistant](https://mlsanity.com/paper/2509.15237/) (2025)

## Built on by (most cited first, 8 of 8)

- [RoboRouter: Training-Free Policy Routing for Robotic Manipulation](https://mlsanity.com/paper/2603.07892/) (2026)
- [Once-For-All: A Train-Once and Select-Anytime Framework for Multimodal Instruction Tuning](https://mlsanity.com/paper/2605.26761/) (2026)
- [Visual Credit Audit for Multimodal Spatial Reasoning](https://mlsanity.com/paper/2607.27069/) (2026)
- [Dive Into the Implicit Biases of Low-rank Vision-language Alignment](https://mlsanity.com/paper/2607.08194/) (2026)
- [Multi-scale Decomposed Convolution Refinement Network for Visible-Infrared Person Re-Identification](https://mlsanity.com/paper/2608.16015/) (2026)
- [Same Reward, Different Skills: When Multimodal RL Learns to Look](https://mlsanity.com/paper/2610.01908/) (2026)
- [EgoOmniWorld: Unifying Egocentric Understanding via Ego-Action-World Interactions](https://mlsanity.com/paper/2610.18499/) (2026)
- [More Than Seeing: Vision as Scalable Anchors in Native Multimodal Pre-Training](https://mlsanity.com/paper/2610.19443/) (2026)

## Abstract (arXiv)

The effectiveness of multimodal instruction tuning depends not only on dataset scale, but critically on whether training samples genuinely require visual reasoning. However, existing instruction datasets often contain a substantial portion of visually redundant samples (solvable from text alone), as well as multimodally misaligned supervision that can degrade learning. To address this, we propose VisNec (Visual Necessity Score), a principled data selection framework that measures the marginal contribution of visual input during instruction tuning. By comparing predictive loss with and without visual context, VisNec identifies whether a training instance is vision-critical, redundant, or misaligned. To preserve task diversity, we combine VisNec with semantic clustering and select high-necessity samples within each cluster. Across 10 downstream benchmarks, training on only 15% of the LLaVA-665K dataset selected by VisNec achieves 100.2% of full-data performance. On the smaller Vision-Flan-186K dataset, our selection not only further reduces data size but also surpasses full-data training by 15.8%. These results demonstrate that measuring and leveraging visual necessity provides an effective solution for both efficient and robust multimodal instruction tuning. Codes and selected subsets will be released upon acceptance.