Parameter-Efficient Adaptation for MLLMs via Implicit Modality Decomposition
Open paper
Benchmark
Model
Rank
Results
audio-visual-question-answering-on-music-avqa
IMoD (Qwen3-8B) (extra data)
#2
Acc: 80.13
visual-question-answering-vqa-on-seed-seedbench
IMoD (LLaMA3) (extra data)
#31
Accuracy: 57.26
Rank counts only results with a code link.