从位置到功能：多模态大语言模型中的空间-功能智能基准
From Where Things Are to What They Are For: Benchmarking Spatial-Functional Intelligence in Multimodal LLMs
该基准为视觉语言模型提供了诊断空间记忆与功能推理结合能力的评测工具，可用于检验模型在具身场景下的高阶认知瓶颈，并指导融合感知、记忆与推理的模型设计。
原文依据 we introduce the Spatial-Functional Intelligence Benchmark (SFI-Bench), a video-based benchmark with over 1,500 expert-annotated questions
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论人类级别的智能体智能超越了低层次的几何感知，从识别物体在哪里演进到理解它们用于什么。尽管现有基准有效评估了多模态大语言模型（MLLMs）的几何感知能力，但它们未能探测实现具身智能所需的高阶认知能力。为填补这一空白，我们引入了空间-功能智能基准（SFI-Bench），这是一个基于视频的基准，包含超过1,500个专家标注的问题，源自多样化的第一人称室内视频扫描。SFI-Bench系统评估高级推理的两个互补维度：（1）结构化空间推理，要求理解复杂布局并形成连贯的空间表征；（2）功能推理，涉及推断物体可供性及其上下文相关的效用。该基准包括条件计数、多跳关系推理、功能配对和基于知识的故障排除等任务，直接挑战模型整合感知、记忆和推理的能力。我们的实验表明，当前MLLMs在结合空间记忆与功能推理及外部知识方面持续挣扎，突显了实现具身智能的关键瓶颈。因此，SFI-Bench提供了一个诊断工具，用于衡量向更具认知能力且真正具身的多模态智能体迈进的进展。
英文原文摘要 1501 字符
Human-level agentic intelligence extends beyond low-level geometric perception, evolving from recognizing where things are to understanding what they are for. While existing benchmarks effectively evaluate the geometric perception capabilities of multimodal large language models (MLLMs), they fall short of probing the higher-order cognitive abilities required for grounded intelligence. To address this gap, we introduce the Spatial-Functional Intelligence Benchmark (SFI-Bench), a video-based benchmark with over 1,500 expert-annotated questions derived from diverse egocentric indoor video scans. SFI-Bench systematically evaluates two complementary dimensions of advanced reasoning: (1) Structured Spatial Reasoning, which requires understanding complex layouts and forming coherent spatial representations, and (2) Functional Reasoning, which involves inferring object affordances and their context-dependent utility. The benchmark includes tasks such as conditional counting, multi-hop relational reasoning, functional pairing, and knowledge-grounded troubleshooting, directly challenging models to integrate perception, memory, and inference. Our experiments reveal that current MLLMs consistently struggle to combine spatial memory with functional reasoning and external knowledge, highlighting a critical bottleneck in achieving grounded intelligence. SFI-Bench therefore provides a diagnostic tool for measuring progress toward more cognitively capable and truly grounded multimodal agents.