AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

SMM Transformer: Leveraging Spiking Neural Networks for Multimodal Tasks

Xiubo Liang, Jinxing Han, Yuke Li, Haoqi Zhu, Yu Zhao, Hongzhi Wang

ICML 2026 · Architectures and efficiency › Spiking networks · approach: Spiking transformers · 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 PLMP neuron (instead of standard LIF with STBP) (ablation)SMM Transformer · NLVR2 devaccuracy: 76.74 -> 78.36Improved
add SMSA (instead of SSA baseline) (ablation)SMM Transformer · NLVR2 devaccuracy: 77.45 -> 78.36Improved
add SMoE modality-aware experts (ablation)SMM Transformer · NLVR2 devaccuracy: 77.55 -> 78.36Improved
add MLM pre-training task to ITC (ablation)SMM Transformer · NLVR2 devaccuracy: 54.74 -> 72.53Improved
add ITM pre-training task to ITC (ablation)SMM Transformer · NLVR2 devaccuracy: 54.74 -> 73.63Improved
make LIF parameters learnable (PLMP K=1) (ablation)static LIF · MSCOCO captioningCIDEr: 125.26 -> 125.69Improved
increase branches from K=1 to K=3 (ablation)PLMP · MSCOCO captioningCIDEr: 125.69 -> 127.72Improved
increase branches from K=3 to K=4 (ablation)PLMP · MSCOCO captioningCIDEr: 127.72 -> 127.63No clear effect
multi-stage instead of shared alpha initialization (ablation)PLMP K=3 · MSCOCO captioningCIDEr: 127.72 -> 126.81Mixed
add self-compensation branch to SMSA (ablation)SMSA · NLVR2accuracy: 76.97 -> 77.53Improved
use shared SMSA (with SMoE) instead of separate SMSA for image and text (ablation)SMM Transformer · NLVR2 devaccuracy: 77.53 -> 79.17Improved
replace dense softmax self-attention with SMSA (ablation)SMM Transformer · NLVR2accuracy: 77.80 -> 77.53Hurt

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