MODE: Modality-Decomposed Expert-Level Mixed-Precision Quantization for MoE Multimodal LLMs
About
Mixture-of-Experts Multimodal Large Language Models (MoE-MLLMs) offer remarkable performance but incur prohibitive GPU memory costs, making compression essential. Among PTQ methods, expert-level mixed-precision quantization has proven effective for MoE-LLMs, yet suffers notable degradation on MoE-MLLMs due to two overlooked biases in expert importance estimation. (1) At the cross-modal level, the numerical dominance of vision tokens causes expert selection frequency to be dominated by vision tokens, masking experts that are critical to the text modality; (2) at the intra-vision level, the large proportion of redundant vision tokens further skew frequency statistics, obscuring experts critical for informative visual content. To bridge gaps, we propose MODE, a modality-decomposed expert-level mixed-precision quantization framework for MoE-MLLMs that decomposes expert selection frequency by modality, filters redundant vision tokens to obtain denoised visual frequency, and further evaluates quantization sensitivity per modality as a complementary signal to frequency-based estimation. These signals are integrated into an Integer Linear Programming formulation to assign per-expert bit-widths under a given budget. Extensive experiments show that MODE is particularly well-suited for MoE-MLLMs, limiting average performance loss to within 2.9% at W3A16, with larger gains at the extreme 2-bit setting.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Multimodal Understanding | MMStar | Accuracy55.68 | 511 | |
| Object Hallucination Evaluation | POPE | Accuracy89.44 | 259 | |
| Visual Question Answering | VizWiz | Accuracy69.42 | 193 | |
| Chart Question Answering | ChartQA | Accuracy87.26 | 165 | |
| Information Visual Question Answering | InfoVQA | Accuracy79.87 | 159 | |
| Text-based Visual Question Answering | TextVQA | TextVQA Accuracy86.24 | 141 | |
| Multimodal Understanding | MMBench | Accuracy82.56 | 137 | |
| Visual Question Answering | GQA | GQA Score60.15 | 75 | |
| Multimodal Understanding | MMMU | Accuracy (MMMU)57.33 | 73 | |
| Multimodal Evaluation | MME-R | Score50.77 | 30 |