AI Paper Atlas

Research map / Architectures and efficiency / Spiking networks

Dance of SNN and ANN: Solving binding problem by combining spike timing and reconstructive attention

Hao Zheng, Hui Lin, Rong Zhao, Luping Shi

NeurIPS 2022 · Architectures and efficiency › Spiking networks · approach: Spiking transformers · 1 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 spike timing dynamics and reconstructive attention instead of using a non-spiking layer (ablation)DASBE · BarsAMI score (%): 9.6±0.04 -> 96.5 ± 0.03Improved
add spike timing dynamics and reconstructive attention instead of using a non-spiking layer (ablation)DASBE · ShapesAMI score (%): 47.5±0.04 -> 90.2 ± 0.07Improved
add spike timing dynamics and reconstructive attention instead of using a non-spiking layer (ablation)DASBE · CornersAMI score (%): 54.2±0.15 -> 80.6 ± 0.08Improved
add spike timing dynamics and reconstructive attention instead of using a non-spiking layer (ablation)DASBE · Multi-MNISTAMI score (%): 43.0± 0.08 -> 55.5 ± 0.06Improved
add spike timing dynamics and reconstructive attention instead of using a non-spiking layer (ablation)DASBE · MNIST+ShapeAMI score (%): 38.3±0.03 -> 53.3 ± 0.26Improved
add a trainable ANN (DAE) instead of using fixed recurrent connections (ablation)DASBE · BarsAMI score (%): 9.3±0.04 -> 96.5 ± 0.03Improved
add a trainable ANN (DAE) instead of using fixed recurrent connections (ablation)DASBE · ShapesAMI score (%): 44.7±0.15 -> 90.2 ± 0.07Improved
add a trainable ANN (DAE) instead of using fixed recurrent connections (ablation)DASBE · CornersAMI score (%): 28.5±0.08 -> 80.6 ± 0.08Improved
add a trainable ANN (DAE) instead of using fixed recurrent connections (ablation)DASBE · Multi-MNISTAMI score (%): 23.0 ± 0.11 -> 55.5 ± 0.06Improved
add a trainable ANN (DAE) instead of using fixed recurrent connections (ablation)DASBE · MNIST+ShapeAMI score (%): 12.9±0.08 -> 53.3 ± 0.26Improved

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