VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis

Hugging Face Daily Papers Papers
spatio-temporal-reasoning multimodal-llm video-benchmark generative-models evaluation active-synthesis benchmark

Summary

VGenST-Bench is a benchmark that uses generative models to actively synthesize controlled spatio-temporal reasoning scenarios, with a multi-agent pipeline and human quality control, to evaluate multimodal large language models.

Spatio-temporal reasoning is a core capability for Multimodal Large Language Models (MLLMs) operating in the real world. As such, evaluating it precisely has become an essential challenge. However, existing spatio-temporal reasoning benchmark datasets primarily rely on static image sets or passively curated video data, which limits the evaluation of fine-grained reasoning capabilities. In this paper, we introduce VGenST-Bench, a video benchmark that employs generative models to actively synthesize highly controlled and diverse evaluation scenarios. To construct VGenST-Bench, we propose a multi-agent pipeline incorporating a human quality control stage, ensuring the quality of all generated videos and QA pairs. We establish a comprehensive 3x2x2 video taxonomy, encompassing Spatial Scale, Perspective, and Scene Dynamics to span diverse scenarios. Furthermore, we design a hierarchical task suite that decouples low-level visual perception from high-level spatio-temporal reasoning. By shifting the paradigm from passive curation to active synthesis, VGenST-Bench enables fine-grained diagnosis of spatio-temporal understanding in MLLMs.
Original Article
View Cached Full Text

Cached at: 05/25/26, 06:36 AM

Paper page - VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis

Source: https://huggingface.co/papers/2605.22570

Abstract

VGenST-Bench presents a video benchmark using generative models for active synthesis of controlled spatio-temporal reasoning scenarios with human quality control.

Spatio-temporal reasoningis a core capability forMultimodal Large Language Models(MLLMs) operating in the real world. As such, evaluating it precisely has become an essential challenge. However, existingspatio-temporal reasoningbenchmark datasets primarily rely on static image sets or passively curated video data, which limits the evaluation of fine-grained reasoning capabilities. In this paper, we introduce VGenST-Bench, avideo benchmarkthat employsgenerative modelsto actively synthesize highly controlled and diverse evaluation scenarios. To construct VGenST-Bench, we propose amulti-agent pipelineincorporating a human quality control stage, ensuring the quality of all generated videos and QA pairs. We establish a comprehensive 3x2x2video taxonomy, encompassing Spatial Scale, Perspective, and Scene Dynamics to span diverse scenarios. Furthermore, we design ahierarchical task suitethat decouples low-level visual perception from high-levelspatio-temporal reasoning. By shifting the paradigm from passive curation to active synthesis, VGenST-Bench enables fine-grained diagnosis of spatio-temporal understanding in MLLMs.

View arXiv pageView PDFProject pageGitHub4Add to collection

Get this paper in your agent:

hf papers read 2605\.22570

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.22570 in a model README.md to link it from this page.

Datasets citing this paper1

#### zino1/VGenST-Bench Viewer• Updated3 days ago • 11k • 35

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.22570 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

A Very Big Video Reasoning Suite

Papers with Code Trending

This paper introduces the Very Big Video Reasoning (VBVR) dataset and benchmark, a large-scale resource with over one million video clips across 200 reasoning tasks, enabling systematic study of spatiotemporal reasoning and showing early signs of emergent generalization.

SVI-Bench: A Dynamic Microworld for Strategic Video Intelligence

Hugging Face Daily Papers

Introduces SVI-Bench, a large-scale benchmark for strategic video intelligence using team sports, designed to evaluate models on dynamic scene understanding, causal reasoning, strategic simulation, and agentic synthesis. The benchmark reveals a capability cliff where models perform well on perceptual tasks but sharply degrade on higher-level strategic reasoning.

Exploring Spatial Intelligence from a Generative Perspective

Hugging Face Daily Papers

Researchers introduce GSI-Bench, the first benchmark to quantify generative spatial intelligence in multimodal models by evaluating 3D spatial constraint compliance during image generation. Fine-tuning on their synthetic dataset boosts both spatial editing fidelity and downstream spatial understanding, showing generative training can strengthen spatial reasoning.

OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMs

Papers with Code Trending

OVO-S-Bench introduces a comprehensive human-annotated benchmark of 1,680 questions across 348 videos to evaluate streaming spatial intelligence in multimodal LLMs, revealing that even the best model (Gemini-3.1-Pro) trails human experts by 27 points. The benchmark exposes key limitations including allocentric mapping as a major bottleneck and chain-of-thought reasoning amplifying spatial errors.