MM-IFEval-Pro: A Multilingual and Attack-Resistant Benchmark for Instruction-Following in Vision-Language Models

SummaryAI generated

On September 7, 2026, the arXiv cs.AI preprint platform released an evaluation benchmark named MM-IFEval-Pro. This benchmark aims to provide a command-following test standard for visual-language models that includes multilingual capabilities and anti-attack features.

Related eventsRELATED EVENTS

2026-09-07 21:02The `GPT-6 Astra model was released, and OpenAI President Brockman confidently declared, “Welcome to the era of AGI.”
2026-09-08 19:21OpenAI released GPT-6 Astra, marking the arrival of the AGI era, with capabilities for autonomous computer operation and scientific research.
2026-09-08 22:40① Nvidia CEO Jensen Huang posted that OpenAI’s GPT-6 Astra, released last week, was trained using approximately 100,000 NV Link 72 clusters, and he believes that General Artificial Intelligence (AGI) has officially arrived; ② GPT-6 Astra can directly operate computers and software to perform complex tasks such as programming, reaching the most advanced level in multiple fields. OpenAI has announced the beginning of the AGI era.

Key entitiesKEY ENTITIES

MM-IFEval-Pro

SignalsSIGNALS

Keyword heat

MM-IFEval-Pro1

All reports (1)SOURCES

↗

A arXiv cs.AI en 2026-09-07 12:00

arXiv:2609.04859v1 发布 MM-IFEval-Pro，这是一个覆盖中文和英文任务及多种指令劫持案例的多模态指令遵循基准。该基准包含 4 大类、24 个子类及 8 类共 52 个子类，每个样本平均包含 3.0 个约束以模拟复杂场景。研究团队构建了包含中文和对抗性指令的强化学习训练集，显著提升了模型在 MM-IFEval-Pro 上的表现，并有效迁移至其他主流多模态基准，展现出强大的跨任务和跨语言泛化能力。