AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

TEFormer: Structured Bidirectional Temporal Enhancement Modeling in Spiking Transformers

Sicheng Shen, Mingyang Lv, Bing Han, Dongcheng Zhao, Guobin Shen, Feifei Zhao, Yi Zeng

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 T-MLP and TEA together (bidirectional temporal fusion) (ablation)H-Spikformer baseline · CIFAR10 (4-384)Acc@1 (%): 95.91 -> 96.24Improved
add T-MLP alone (ablation)H-Spikformer baseline · CIFAR10 (4-384)Acc@1 (%): 95.91 -> 95.98No clear effect
add TEA alone (ablation)H-Spikformer baseline · CIFAR10 (4-384)Acc@1 (%): 95.91 -> 95.85Hurt
assign forward fusion to TEA and backward to T-MLP instead of both backward (ablation)TEFormer · CIFAR10 (4-384)Acc@1 (%): 96.09 -> 96.24Improved
use single-gate T-MLP instead of GRU-style gate (ablation)TEFormer · CIFAR10Acc@1 (%): 95.90 -> 96.24Improved
use single-gate T-MLP instead of EMA-style update (ablation)TEFormer · CIFAR100Acc@1 (%): 79.55 -> 79.84Improved
use learnable alpha instead of fixed 0.75 (ablation)TEA · CIFAR10Acc@1 (%): 96.04 -> 96.24Improved
use learnable alpha instead of fixed 0.75 (ablation)TEA · CIFAR10-DVSAcc@1 (%): 81.2 -> 81.9Improved
increase temporally enhanced layers from 2 to 4 (ablation)TEFormer · CIFAR10Acc@1 (%): 96.04 -> 95.66Hurt
enhance 2 layers instead of 1 (ablation)TEFormer · CIFAR10Acc@1 (%): 95.24 -> 96.04Improved
reduce T-MLP MLP ratio from 4 to 2 (ablation)TEFormer · CIFAR10Acc@1 (%) / energy: 96.24 -> 95.98; 5.18 mJ -> 2.39 mJMixed
add TEFormer under phase encoding vs QKFormerTEFormer · CIFAR10 (phase encoding)Acc@1 (%): 87.76 -> 89.92Improved

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