dorsal/arxiv
View SchemaBabyVision: Visual Reasoning Beyond Language
| Authors | Liang Chen, Weichu Xie, Yiyan Liang, Hongfeng He, Hans Zhao, Zhibo Yang, Zhiqi Huang, Haoning Wu, Haoyu Lu, Y. charles, Yiping Bao, Yuantao Fan, Guopeng Li, Haiyang Shen, Xuanzhong Chen, Wendong Xu, Shuzheng Si, Zefan Cai, Wenhao Chai, Ziqi Huang, Fangfu Liu, Tianyu Liu, Baobao Chang, Xiaobo Hu, Kaiyuan Chen, Yixin Ren, Yang Liu, Yuan Gong, Kuan Li |
|---|---|
| Categories | |
| ArXiv ID | 2601.06521vv1 |
| URL | https://arxiv.org/abs/2601.06521 |
| License | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
Abstract
While humans develop core visual skills long before acquiring language, contemporary Multimodal LLMs (MLLMs) still rely heavily on linguistic priors to compensate for their fragile visual understanding. We uncovered a crucial fact: state-of-the-art MLLMs consistently fail on basic visual tasks that humans, even 3-year-olds, can solve effortlessly. To systematically investigate this gap, we introduce BabyVision, a benchmark designed to assess core visual abilities independent of linguistic knowledge for MLLMs. BabyVision spans a wide range of tasks, with 388 items divided into 22 subclasses across four key categories. Empirical results and human evaluation reveal that leading MLLMs perform significantly below human baselines. Gemini3-Pro-Preview scores 49.7, lagging behind 6-year-old humans and falling well behind the average adult score of 94.1. These results show despite excelling in knowledge-heavy evaluations, current MLLMs still lack fundamental visual primitives. Progress in BabyVision represents a step toward human-level visual perception and reasoning capabilities. We also explore solving visual reasoning with generation models by proposing BabyVision-Gen and automatic evaluation toolkit. Our code and benchmark data are released at https://github.com/UniPat-AI/BabyVision for reproduction.
{
"annotation_id": "6e0655ac-c695-47bf-86d2-6ac4f3fc454d",
"date_created": "2026-02-17T05:53:08.045000Z",
"date_modified": "2026-02-17T05:53:08.045000Z",
"file_hash": "f384cd56f8bef91557775dacadab2e146daa9b7dafc3124191cc621f7e4cdc0a",
"private": false,
"record": {
"abstract": "While humans develop core visual skills long before acquiring language, contemporary Multimodal LLMs (MLLMs) still rely heavily on linguistic priors to compensate for their fragile visual understanding. We uncovered a crucial fact: state-of-the-art MLLMs consistently fail on basic visual tasks that humans, even 3-year-olds, can solve effortlessly. To systematically investigate this gap, we introduce BabyVision, a benchmark designed to assess core visual abilities independent of linguistic knowledge for MLLMs. BabyVision spans a wide range of tasks, with 388 items divided into 22 subclasses across four key categories. Empirical results and human evaluation reveal that leading MLLMs perform significantly below human baselines. Gemini3-Pro-Preview scores 49.7, lagging behind 6-year-old humans and falling well behind the average adult score of 94.1. These results show despite excelling in knowledge-heavy evaluations, current MLLMs still lack fundamental visual primitives. Progress in BabyVision represents a step toward human-level visual perception and reasoning capabilities. We also explore solving visual reasoning with generation models by proposing BabyVision-Gen and automatic evaluation toolkit. Our code and benchmark data are released at https://github.com/UniPat-AI/BabyVision for reproduction.",
"arxiv_id": "2601.06521",
"authors": [
"Liang Chen",
"Weichu Xie",
"Yiyan Liang",
"Hongfeng He",
"Hans Zhao",
"Zhibo Yang",
"Zhiqi Huang",
"Haoning Wu",
"Haoyu Lu",
"Y. charles",
"Yiping Bao",
"Yuantao Fan",
"Guopeng Li",
"Haiyang Shen",
"Xuanzhong Chen",
"Wendong Xu",
"Shuzheng Si",
"Zefan Cai",
"Wenhao Chai",
"Ziqi Huang",
"Fangfu Liu",
"Tianyu Liu",
"Baobao Chang",
"Xiaobo Hu",
"Kaiyuan Chen",
"Yixin Ren",
"Yang Liu",
"Yuan Gong",
"Kuan Li"
],
"categories": [
"cs.CV",
"cs.CL"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"title": "BabyVision: Visual Reasoning Beyond Language",
"url": "https://arxiv.org/abs/2601.06521",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "2a0c9ce8-df10-49e7-b8b6-bf047f8b1c3f",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}