AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

VEQ: Modality-Adaptive Quantization for MoE Vision-Language Models

Guangshuo Qin, Zhiteng Li, Zheng Chen, Weihang Zhang, Linghe Kong, Yulun Zhang

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multi-modality image fusion · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add gradient scaling factor gamma to expert importance (ablation)VEQ-ME · MMMU/InfoVQA/ScienceQA (Kimi-VL W3)average accuracy (%): 62.02 -> 63.27Improved
add quantity normalization factor beta to expert importance (ablation)VEQ-ME · MMMU/InfoVQA/ScienceQA (Kimi-VL W3)average accuracy (%): 62.30 -> 63.27Improved
add router affinity p to Hessian weighting (ablation)VEQ-MA · MMMU/InfoVQA/ScienceQA (Kimi-VL W3)average accuracy (%): 63.63 -> 63.96Improved
add modality indicator alpha to Hessian weighting (ablation)VEQ-MA · MMMU/InfoVQA/ScienceQA (Kimi-VL W3)average accuracy (%): 62.46 -> 63.96Improved
replace GPTQ with VEQ-MA at W3GPTQ · Kimi-VL average of 9 benchmarks (W3)average accuracy (%): 62.33 -> 65.41Improved
replace AWQ with VEQ-ME at W3AWQ · Kimi-VL average of 9 benchmarks (W3)average accuracy (%): 63.37 -> 64.00Improved
replace GPTQ with VEQ-MA at W3GPTQ · Qwen3-VL average of 9 benchmarks (W3)average accuracy (%): 64.16 -> 67.14Improved
replace GPTQ with VEQ-MA at W4GPTQ · Qwen3-VL average of 9 benchmarks (W4)average accuracy (%): 73.20 -> 74.36Improved
replace GPTQ with VEQ-MA at W4GPTQ · Kimi-VL average of 9 benchmarks (W4)average accuracy (%): 72.96 -> 73.57Improved
replace AWQ with VEQ-ME at W4AWQ · Kimi-VL average of 9 benchmarks (W4)average accuracy (%): 72.63 -> 72.86Improved
replace AWQ with VEQ-ME at W3AWQ · Qwen3-VL average of 9 benchmarks (W3)average accuracy (%): 62.22 -> 63.93Improved
proportionally scale gamma and beta together (ablation)VEQ-ME · validation PPL (64 MMMU samples)PPLNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy