AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

SAQ: Stabilizer-Aware Quantum Error Correction Decoder

David Zenati, Eliya Nachmani

ICLR 2026 · Architectures and efficiency › RNNs and normalization · approach: Scalable approximation and equivariant compute · 3 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 logical classification loss (ablation)SAQ-Decoder · Toric code depolarizing, L=6average LER: 2.113e-01 -> 1.972e-01Improved
add logical prior loss (ablation)SAQ-Decoder · Toric code depolarizing, L=6average LER: 2.055e-01 -> 1.972e-01Improved
add entropy regularization loss (ablation)SAQ-Decoder · Toric code depolarizing, L=6average LER: 2.047e-01 -> 1.972e-01Improved
add global token (ablation)SAQ-Decoder · Toric code depolarizing, L=6average LER: ~0.21 -> ~0.19Improved
add attention masking (ablation)SAQ-Decoder · Toric code depolarizing, L=6LERImproved
use asymmetric syndrome-to-logical cross-attention instead of symmetric (ablation)SAQ-Decoder · Toric code depolarizing, L=6LERImproved
use both syndrome and logical streams instead of one stream (ablation)SAQ-Decoder · Toric code depolarizing, L=6LERImproved
share weights across layers instead of separate weights per stream (ablation)SAQ-Decoder · Toric code depolarizing, L=6LER and parametersMixed
add CPND post-processing refinement to the neural decoder (ablation)SAQ-Decoder · Topological codesLERNo clear effect
use SAQ-Decoder instead of QECCTQECCT · Toric code depolarizingerror threshold: 17.8% -> 18.6%Improved
use SAQ-Decoder instead of QECCTQECCT · Rotated surface code depolarizingerror threshold: 17.2% -> 18.3%Improved
use SAQ-Decoder instead of QECCTQECCT · Toric code L=10 depolarizinginference time (ms): 2.3 -> 0.45Improved

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