AI RESEARCH
Chain of Modality: From Static Fusion to Dynamic Orchestration in Omni-MLLMs
arXiv CS.CV
•
ArXi:2604.14520v1 Announce Type: new Omni-modal Large Language Models (Omni-MLLMs) promise a unified integration of diverse sensory streams. However, recent evaluations reveal a critical performance paradox: unimodal baselines frequently outperform joint multimodal inference. We trace this perceptual fragility to the static fusion topologies universally employed by current models, identifying two structural pathologies: positional bias in sequential inputs and alignment traps in interleaved formats, which systematically distort attention regardless of task semantics.