NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs
Quick summary
arXiv:2608.08107v1 Announce Type: cross Abstract: Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining. In this work, we investigate this phenomenon from the perspective of internal adaptation dynamics and discover that neurons in pretrained LLMs exhibit heterogeneous plasticity during multimodal learning: some neurons are critical for preserving language capabilities, while others are more adaptive to multimodal knowledge. Based on this insight, we propose NeuPAT (Neuron-aware Plast
Key takeaways
- arXiv:2608.08107v1 Announce Type: cross Abstract: Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining.
- In this work, we investigate this phenomenon from the perspective of internal adaptation dynamics and discover that neurons in pretrained LLMs exhibit heterogeneous plasticity during multimodal learning: some neurons are critical for preserving language capabilities, while others are more adaptive to multimodal knowledge.
- Based on this insight, we propose NeuPAT (Neuron-aware Plast
Why it matters
“NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs” should be evaluated beyond branding and benchmark scores. Its practical importance will emerge in task accuracy, latency, unit cost, safety and integration with real workflows.
Member comments