Research Papers 论文研究 1d ago Updated 20h ago 更新于 20小时前 48

Compliance, Capability, and Conflict: Benchmarking Multimodal LLMs under System Messages 合规、能力与冲突:系统消息下多模态大语言模型的基准测试

Introduces VSysBench, a benchmark built on MMVet-v2 that evaluates how Multimodal Large Language Models (MLLMs) adhere to system messages across 5 main categories and 22 sub-categories of constraints Proposes two evaluation metrics: Joint Satisfaction Rate (JSR) measuring simultaneous constraint compliance and answer correctness, and Cross-Constraint Sensitivity (CCS) assessing how constraints interact System messages substantially erode base task accuracy across 16 evaluated MLLMs, revealing a 提出VSysBench基准测试,首次系统评估多模态LLM在系统消息下的合规性、能力和冲突表现 将约束分为5大类22个子类,涵盖从文本指令到视觉接地指令的完整范围,每类配对抗样本测试指令层级 引入联合满意度率(JSR)和跨约束敏感性(CCS)两个评估指标,联合衡量约束合规性与答案正确性 测试16个MLLMs发现:系统消息显著侵蚀基础任务准确率;开放权重模型在用户冲突下合规性崩溃,顶级专有模型保持稳定;视觉接地约束对所有模型都是最难类别

65
Hot 热度
72
Quality 质量
68
Impact 影响力

Analysis 深度分析

TL;DR

  • Introduces VSysBench, a benchmark built on MMVet-v2 that evaluates how Multimodal Large Language Models (MLLMs) adhere to system messages across 5 main categories and 22 sub-categories of constraints
  • Proposes two evaluation metrics: Joint Satisfaction Rate (JSR) measuring simultaneous constraint compliance and answer correctness, and Cross-Constraint Sensitivity (CCS) assessing how constraints interact
  • System messages substantially erode base task accuracy across 16 evaluated MLLMs, revealing a significant compliance-capability tradeoff
  • Open-weight models show compliance collapse under user conflict, while top proprietary models maintain stable compliance
  • Vision-grounded constraints are consistently the hardest category for every model tested

Why It Matters

This research addresses a critical gap in MLLM evaluation: existing benchmarks fail to measure how well models follow system-level instructions in multimodal contexts, despite production deployments increasingly relying on them. The findings have direct implications for practitioners deploying MLLMs in real-world scenarios where system messages must coexist with strong vision-language capabilities, and where user conflicts may arise between system directives and user prompts.

Technical Details

  • VSysBench is constructed on top of MMVet-v2, organizing constraints into 5 main categories and 22 sub-categories spanning from textual directives in visual contexts to fully vision-grounded constraints, each paired with a misaligned counterpart to stress-test the instructional hierarchy
  • Two novel metrics are introduced: Joint Satisfaction Rate (JSR) scores responses jointly along constraint compliance and answer correctness axes, while Cross-Constraint Sensitivity (CCS) measures how the presence of one constraint affects compliance with another
  • The benchmark evaluates 16 MLLMs (both open-weight and proprietary), testing them under conditions where system messages may conflict with user instructions
  • Constraints range across textual directives in visual contexts to fully vision-grounded ones, with misaligned counterparts designed to create tension between system and user instructions

Industry Insight

  • Deployment pipelines for MLLMs must account for the significant accuracy degradation caused by system messages; practitioners should expect and plan for capability tradeoffs when implementing behavioral constraints
  • Open-weight models remain vulnerable to instruction hierarchy breakdown under user conflict, making proprietary models a safer choice for production environments where adversarial or conflicting user inputs are likely
  • Vision-grounded constraints represent the hardest evaluation category across all models, suggesting that future research should prioritize improving how models integrate visual information with system-level behavioral instructions

TL;DR

  • 提出VSysBench基准测试,首次系统评估多模态LLM在系统消息下的合规性、能力和冲突表现
  • 将约束分为5大类22个子类,涵盖从文本指令到视觉接地指令的完整范围,每类配对抗样本测试指令层级
  • 引入联合满意度率(JSR)和跨约束敏感性(CCS)两个评估指标,联合衡量约束合规性与答案正确性
  • 测试16个MLLMs发现:系统消息显著侵蚀基础任务准确率;开放权重模型在用户冲突下合规性崩溃,顶级专有模型保持稳定;视觉接地约束对所有模型都是最难类别

为什么值得看

本文填补了多模态场景下系统消息合规性评估的空白,揭示了约束遵循与基础能力之间的权衡关系,为生产部署中的MLLM安全与可靠性提供了量化评估工具。

技术解析

  • VSysBench基准:基于MMVet-v2构建,组织约束为5个主类别和22个子类别,从视觉上下文中的文本指令到完全视觉接地指令,每类配对不匹配 counterpart 以压力测试指令层级
  • 评估指标:联合满意度率(JSR)联合评估约束合规性和答案正确性;跨约束敏感性(CCS)衡量约束间的相互影响
  • 实验规模:测试16个多模态大语言模型,涵盖开放权重和专有模型
  • 关键发现:系统消息引入后基础任务准确率显著下降;开放权重模型在用户冲突场景下合规性崩溃,而顶级专有模型保持稳定;视觉接地约束是各类模型中最难处理的类别

行业启示

  • 生产部署MLLM时需权衡系统消息的合规控制与基础能力损失,建议建立分层约束策略以最小化能力侵蚀
  • 开放权重模型在对抗性用户冲突下存在安全风险,企业级应用应优先选择经过严格安全对齐的专有模型或加强开源模型的安全微调
  • 视觉接地约束是当前多模态模型的能力瓶颈,未来研究应聚焦于提升模型对视觉上下文约束的理解和执行能力

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Multimodal 多模态 LLM 大模型 Benchmark 基准测试 Evaluation 评测 Alignment 对齐