AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Learning with Differentiable Pertubed Optimizers

Quentin Berthet, Mathieu Blondel, Olivier Teboul, Marco Cuturi, Jean-Philippe Vert, Francis R. Bach

NeurIPS 2020 · Learning theory and algorithms › Prediction and decision losses · approach: Active learning and subset selection · 32 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
use perturbed Fenchel-Young loss instead of blackbox lossaffine model · Label ranking, 21 datasetsfraction of datasets better or similar (within 5%): 76% better, 90% better or similarImproved
use perturbed Fenchel-Young loss instead of blackbox lossaffine model · Label ranking: authorshipSpearman correlation: 0.76 -> 0.95Improved
use Fenchel-Young loss instead of squared lossaffine model · Label ranking: housingSpearman correlation: 0.70 -> 0.75Improved
use Fenchel-Young loss instead of squared lossaffine model · Label ranking: authorshipSpearman correlation: 0.96 -> 0.95No clear effect
use Fenchel-Young loss instead of squared loss on perturbed argmax (ablation)affine model · Label ranking: authorshipSpearman correlation: 0.28 -> 0.95Improved
also tune temperature epsilon for perturbed+squared loss (ablation)affine model · Label ranking: authorshipSpearman correlation: 0.28 -> 0.93Improved
very high or very low temperature epsilon (ablation)perturbed FY · CIFAR-10 vanilla-CNNtrain/test squared lossHurt
perturbed FY loss vs cross entropyvanilla CNN · CIFAR-10test accuracyNo clear effect
perturbed FY with single perturbation M=1 vs blackbox/squaredResNet18 first 5 layers · Warcraft shortest pathperfect accuracy / cost ratioImproved

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