Research Papers 论文研究 5h ago Updated 1h ago 更新于 1小时前 48

Compositional Failure in Audio-Visual LLMs: Late-Layer Prior Dominance Under Cross-modal Conflict 音频视觉大语言模型的组合性失败:跨模态冲突下的晚期层先验主导

Audio-visual LLMs exhibit a failure mode called "prior dominance" where they commit to internally preferred answer patterns rather than resolving cross-modal conflicts between audio and video inputs VideoLLaMA 2-7B-AV remains near chance performance on AVHBench's exact-string Yes/No subset across three alignment configurations, despite substantial shifts in output priors InternVideo2 suffers a 32.3% accuracy drop under cross-modal conflict conditions, with an additional 17.3% instruction-followi 研究音频-视觉冲突作为AV-LLMs的组合泛化测试,发现模型在跨模态冲突下表现显著下降 提出"先验主导"(prior dominance)失败模式:晚期层过度依赖内部偏好答案模式,与冲突输入关联较弱 VideoLLaMA 2-7B-AV三种对齐配置在AVHBench上准确率接近随机水平,InternVideo2准确率下降32.3%且指令遵循失败率17.3% 机制可解释性分析显示承诺集中在25.5±1层,更强时间对齐仅改变答案偏见而不改善冲突解决 代码和数据已开源,支持复现机制审计和行为评估

65
Hot 热度
75
Quality 质量
68
Impact 影响力

Analysis 深度分析

TL;DR

  • Audio-visual LLMs exhibit a failure mode called "prior dominance" where they commit to internally preferred answer patterns rather than resolving cross-modal conflicts between audio and video inputs
  • VideoLLaMA 2-7B-AV remains near chance performance on AVHBench's exact-string Yes/No subset across three alignment configurations, despite substantial shifts in output priors
  • InternVideo2 suffers a 32.3% accuracy drop under cross-modal conflict conditions, with an additional 17.3% instruction-following failure rate
  • Mechanistic interpretability analysis reveals that commitment to answers concentrates at layers 25.5 ± 1, indicating late-layer dominance in decision-making
  • Stronger temporal alignment modifies answer bias but fails to improve compositional conflict resolution, suggesting a fundamental architectural limitation

Why It Matters

This research exposes a critical vulnerability in multimodal AI systems that process both audio and video: when sensory inputs conflict, models default to internal priors rather than performing genuine cross-modal reasoning. For practitioners building AV-LLM applications, this means current systems may appear competent on aligned data but fail catastrophically in real-world scenarios where audio-visual mismatches occur, such as in video analysis, robotics, or interactive AI systems.

Technical Details

  • Benchmark and Models: Evaluated on AVHBench using VideoLLaMA 2-7B-AV (three alignment configurations) and off-the-shelf InternVideo2, testing exact-string Yes/No responses under cross-modal conflict conditions
  • Prior Dominance Mechanism: Late-layer commitment (concentrated at 25.5 ± 1 layers) to internally preferred answer patterns that are weakly grounded in the actual conflicting inputs, identified through mechanistic interpretability analysis
  • Temporal Alignment Findings: Strengthening temporal alignment between audio and video streams shifts answer bias but does not resolve compositional conflict, indicating the problem is architectural rather than purely alignment-related
  • Quantified Failures: 32.3% accuracy decrease under cross-modal conflict for InternVideo2, with 17.3% instruction-following failure rate; VideoLLaMA 2-7B-AV remains near chance despite output prior shifts
  • Reproducibility: Code and data for mechanistic audit and behavioral evaluations are publicly available

Industry Insight

  • Developers of multimodal AI systems should treat cross-modal conflict scenarios as essential stress tests, not edge cases, since real-world deployments will inevitably encounter audio-visual mismatches that expose prior dominance failures
  • The finding that late-layer commitment drives this failure suggests architectural interventions—such as intermediate fusion mechanisms or layer-aware conflict resolution—may be more effective than simply improving temporal alignment
  • Researchers should prioritize mechanistic interpretability audits for multimodal models to identify where and how conflicting evidence is (or is not) being integrated before final predictions are made

TL;DR

  • 研究音频-视觉冲突作为AV-LLMs的组合泛化测试,发现模型在跨模态冲突下表现显著下降
  • 提出"先验主导"(prior dominance)失败模式:晚期层过度依赖内部偏好答案模式,与冲突输入关联较弱
  • VideoLLaMA 2-7B-AV三种对齐配置在AVHBench上准确率接近随机水平,InternVideo2准确率下降32.3%且指令遵循失败率17.3%
  • 机制可解释性分析显示承诺集中在25.5±1层,更强时间对齐仅改变答案偏见而不改善冲突解决
  • 代码和数据已开源,支持复现机制审计和行为评估

为什么值得看

本文为多模态大模型在跨模态冲突场景下的系统性缺陷提供了实证证据和机制解释,揭示了当前AV-LLMs组合泛化能力的根本局限。通过机制可解释性分析定位问题根源,为后续改进提供了明确方向,对多模态模型架构设计和评估具有重要参考价值。

技术解析

  • 基准测试:使用AVHBench的精确字符串Yes/No子集,测试模型在同步但语义不兼容的音视频证据下的匹配判断能力
  • 模型评估:VideoLLaMA 2-7B-AV三种对齐配置表现接近随机,InternVideo2在跨模态冲突下准确率下降32.3%,伴随17.3%指令遵循失败
  • 机制可解释性分析:发现模型承诺集中在25.5±1层,表明晚期层过度依赖内部偏好答案模式而非冲突输入
  • 时间对齐实验:更强的时间对齐能改变答案偏见,但不改善组合冲突解决能力,说明问题根源不在对齐精度

行业启示

  • 当前AV-LLMs在跨模态冲突处理上存在系统性缺陷,需从模型架构层面改进组合泛化能力,而非仅优化对齐策略
  • 多模态模型评估应纳入跨模态冲突测试,以检测"先验主导"等隐蔽失败模式,避免在标准基准上高估模型能力
  • 建议研究团队开发专门的冲突解决机制(如早期层干预、对比学习增强),并在训练数据中增加跨模态冲突样本以提升鲁棒性

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。

LLM 大模型 Multimodal 多模态 Evaluation 评测 Benchmark 基准测试 Research 科学研究