提出三视图基准测试，揭示多模态大模型在复杂结构推理中的系统性局限。
TriViewBench: Controlled Complexity Scaling for Multi-View Structural Reasoning in MLLMs
- 构建合成3D场景的三视图基准，精确控制物体数量与遮挡程度。
- 所有模型在复杂度提升时性能下降，全局恢复能力崩溃达80.02%。
- 揭示模型因视角混淆导致误计数，且思维链提示几乎无效。
多模态大语言模型（MLLMs）在标准视觉问答任务中表现优异，但在可控结构复杂度下的可扩展性仍不明确。本文提出TriViewBench，一个基于合成3D场景的三视图视觉推理基准，显式参数化物体数量与遮挡程度。该基准包含1,923个场景和超过14,000个问答对，分为四个复杂度层级和三种推理类型：局部决策、物体计数、全局恢复。我们在统一提示协议下评估18个开源与闭源MLLM。所有模型均表现出一致的能力层级（局部决策 > 物体计数 > 全局恢复），且性能随复杂度单调下降：局部决策仅下降12.11%，物体计数下降59.14%，全局恢复崩溃达80.02%。对物体计数的错误分析显示两种独立失败模式：单视图任务因遮挡盲区导致漏计，多视图任务则因跨视图身份混淆导致多计。思维链（CoT）提示整体效果近乎为零（Δ= -0.16%），且对全局恢复的影响强依赖模型能力，表明瓶颈在于跨视图空间表征而非推理策略。这些发现揭示了当前MLLM在结构推理上的根本局限，并确立TriViewBench作为诊断结构性推理失败的可控框架。
原文摘要 · Abstract (English)
Multimodal Large Language Models (MLLMs) demonstrate strong performance on standard visual question answering benchmarks, yet their scalability under controlled structural complexity remains poorly understood. We introduce TriViewBench, a controlled three-view visual reasoning benchmark constructed from synthetic 3D scenes with explicitly parameterized object count and occlusion. The benchmark contains 1,923 scenes and over 14K Question-Answer (QA) pairs organized into four complexity levels and three reasoning categories: Local Decision, Object Counting, and Global Recovery. We evaluate 18 open- and closed-source MLLMs under a unified prompting protocol. All 18 models exhibit an identical capability hierarchy without exception (Local Decision > Object Counting > Global Recovery), and performance degrades monotonically with complexity: Local Decision tasks decline modestly (12.11% relative drop), while Object Counting degrades substantially (59.14%) and Global Recovery collapses severely (80.02%). Error analysis on Object Counting reveals two mechanistically independent failure modes: single-view tasks are dominated by undercounting due to occlusion blindness, whereas the multi-view task reverses to overcounting due to cross-view identity confusion. Chain-of-Thought (CoT) prompting yields near-zero overall benefit ($Δ= -0.16\%$) and its effect on Global Recovery is strongly capability-gated, suggesting that the bottleneck lies in cross-view spatial representation rather than reasoning strategy. These findings reveal fundamental scalability limitations in current MLLMs and position TriViewBench as a controlled diagnostic framework for analyzing structural reasoning failures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。