PulseAugur
EN
LIVE 09:33:17

New HumanSightBench benchmark reveals MLLMs struggle with visual perception

By PulseAugur Editorial · [1 sources] ·

A new benchmark, HumanSightBench, has been developed to evaluate the visual perception capabilities of multimodal large language models (MLLMs). This benchmark includes 2,876 visual questions designed to be easy for humans but challenging for models, passing through five inclusion gates to minimize confounds like ambiguity and shortcuts. Initial evaluations show that even the top-performing MLLMs struggle with tasks like counting, with scores below 27.5%, highlighting significant gaps in their visual understanding compared to human perception. AI

IMPACT Highlights significant gaps in MLLM visual perception, guiding future training and evaluation efforts.

RANK_REASON The cluster describes a new benchmark for evaluating AI models, which falls under research. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New HumanSightBench benchmark reveals MLLMs struggle with visual perception

How we ranked this

Signal score
16 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The cluster describes a new benchmark for evaluating AI models, which falls under research. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. arXiv cs.AI TIER_1 English(EN) · Zixin Su, Yangguang Lin, Zegang Peng, Yuntong Sun, Xuanyi Wang, Jiahao Wang, Qiankai Xu, Bing Yuan, Jingjia Huang, Ruyang Liu, Yanxu Hu, Yufeng Zhou, Yi Lin ·

    HumanSightBench: A Benchmark for Human Visual Perception in MLLMs

    arXiv:2611.12305v2 Announce Type: replace-cross Abstract: Multimodal large language models (MLLMs) have demonstrated strong performance on knowledge-intensive and reasoning-based vision-language benchmark datasets, but still struggle to solve the visual questions that people can …

RELATED ENTITIES

RELATED TOPICS