AI Paper Atlas

Research map / Probabilistic and scientific modeling / Dynamical systems

Noisy Recurrent Neural Networks

Soon Hoe Lim, N. Benjamin Erichson, Liam Hodgkinson, Michael W. Mahoney

NeurIPS 2021 · Probabilistic and scientific modeling › Dynamical systems · approach: Latent dynamics models · 6 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
noise injection (NRNN 0.02/0.05) vs deterministic Lipschitz RNNLipschitz RNN · Ordered pixel-by-pixel MNIST, white noise σ=0.3test accuracy (%): 47.1 -> 73.5Improved
noise injection (NRNN 0.02/0.05) vs Lipschitz RNNLipschitz RNN · Ordered pixel-by-pixel MNIST, S&P α=0.1test accuracy (%): 73.5 -> 85.5Improved
noise injection (NRNN 0.02/0.05) vs Lipschitz RNNLipschitz RNN · Ordered pixel-by-pixel MNIST, adversarial r=0.15test accuracy (%): 37.1 -> 70.6Improved
noise injection (NRNN 0.02/0.05) vs Lipschitz RNNLipschitz RNN · Ordered pixel-by-pixel MNIST, cleantest accuracy (%): 99.2 -> 99.1Hurt
increase additive noise from 0.02 to 0.05 (multiplicative 0.02)NRNN · Ordered pixel-by-pixel MNIST, white noise σ=0.3test accuracy (%): 62.9 -> 73.5Improved
increase additive noise from 0.02 to 0.05 (multiplicative 0.02)NRNN · Ordered pixel-by-pixel MNIST, adversarial r=0.15test accuracy (%): 58.3 -> 70.6Improved
increase additive noise from 0.02 to 0.05 (multiplicative 0.02)NRNN · Ordered pixel-by-pixel MNIST, cleantest accuracy (%): 99.1 -> 99.1No clear effect
NRNN 0.02/0.02 vs Lipschitz RNNLipschitz RNN · Ordered pixel-by-pixel MNIST, white noise σ=0.2test accuracy (%): 78.9 -> 88.4Improved
NRNN 0.02/0.02 vs CoRNNCoRNN · Ordered pixel-by-pixel MNIST, adversarial r=0.15test accuracy (%): 35.1 -> 58.3Improved
noise injection (NRNN 0.03/0.06) vs Lipschitz RNNLipschitz RNN · ECG, white noise σ=1.2test accuracy (%): 88.9 -> 92.6Improved
noise injection (NRNN 0.03/0.06) vs Lipschitz RNNLipschitz RNN · ECG, multiplicative noise σM=1.2test accuracy (%): 95.6 -> 96.5Improved
noise injection (NRNN 0.03/0.06) vs Lipschitz RNNLipschitz RNN · ECG, cleantest accuracy (%): 97.7 -> 97.7No clear effect

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