AI Paper Atlas

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

Rethinking Gating Mechanism in Sparse MoE: Handling Arbitrary Modality Inputs with Confidence-Guided Gate

Liangwei Nathan Zheng, Wei Emma Zhang, Mingyu Guo, Olaf Maennel, Weitong Chen

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities

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
use ConfNet confidence gating instead of softmax (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 48.59 -> 53.44Improved
use ConfNet confidence gating instead of softmax with load-balance loss (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 51.67 -> 53.44Improved
use ConfNet instead of mean uniform expert weights (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 48.34 -> 53.44Improved
use ConfNet instead of Gaussian gating (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 46.50 -> 53.44Improved
use ConfNet instead of Laplacian gating (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 50.10 -> 53.44Improved
use ConfNet instead of LFB gating (ablation)ConfSMoE · MIMIC-III 48-IHMF1: 48.36 -> 53.44Improved
use ConfNet instead of softmax (ablation)ConfSMoE · CMU-MOSI 50% missingAUC: 68.77 -> 70.41Improved
use ConfNet instead of softmax with load-balance loss (ablation)ConfSMoE · CMU-MOSI 50% missingF1: 42.43 -> 44.34Improved
add load-balance loss to softmax gating (ablation)softmax router · MIMIC-III 48-IHMF1: 48.59 -> 51.67Improved
add load-balance loss to softmax gating (ablation)softmax router · CMU-MOSI 50% missingAUC: 68.77 -> 67.40Hurt
Laplacian and Gaussian gates without load-balance loss vs softmax w/ load-balance (ablation)gating function · MIMIC-IIIF1: 46.50 -> 51.67No clear effect
add the two-stage imputation module (with ConfNet) (ablation)ConfSMoE · MIMIC / CMU datasetsF1 and AUCImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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