AI Paper Atlas

Research map / Optimization / Optimizers

AGD: an Auto-switchable Optimizer using Stepwise Gradient Difference for Preconditioning Matrix

Yun Yue, Zhiling Ye, Jiadi Jiang, Yongchao Liu, Ke Zhang

NeurIPS 2023 · Optimization › Optimizers · approach: Second-order and structured preconditioning · 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
AGD instead of AdaBeliefAdaBelief · PTB 1-layer LSTMtest perplexity: 84.40 -> 81.23Improved
AGD instead of AdaBeliefAdaBelief · PTB 3-layer LSTMtest perplexity: 61.34 -> 60.89Improved
AGD instead of AdaBeliefAdaBelief · IWSLT14 De-EnBLEU: 35.93 -> 35.94No clear effect
AGD instead of AdamAdam · ImageNet ResNet18top-1 accuracy: 64.03 -> 70.35Improved
AGD instead of SGDSGD · ImageNet ResNet18top-1 accuracy: 69.94 -> 70.35Improved
AGD instead of SGDSGD · CIFAR-10 ResNet32top-1 accuracy: 93.10 -> 93.12No clear effect
AGD instead of AdamAdam · Avazu MLPtest AUC: 0.7458 -> 0.7480Improved
AGD instead of AdamAdam · Criteo DCNtest AUC: 0.8023 -> 0.8023No clear effect
AGD instead of SGDSGD · Criteo DCNtest AUC: 0.7296 -> 0.8023Improved
add AMSGrad condition to AGD (ablation)AGD · CIFAR-10 ResNet20top-1 accuracy: 92.35 -> 92.25Hurt
AGD with stepwise-gradient-difference preconditioner instead of AdaHessian (cost)AdaHessian · IWSLT14 Transformer smalltime per epoch: 750 -> 278Improved
AGD instead of AdamW (cost)AdamW · IWSLT14 Transformer smalltime per epoch: 260 -> 278No clear effect

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