AI Paper Atlas

Research map / Optimization / Nonconvex and smooth optimization

ProxQuant: Quantized Neural Networks via Proximal Operators

Yu Bai, Yu-Xiang Wang, Edo Liberty

ICLR 2019 · Optimization › Nonconvex and smooth optimization · approach: Coordinate and Adam convergence analysis · 71 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 ProxQuant-Binary instead of BinaryConnectResNet binary quantization · CIFAR-10top-1 classification error (%): 9.54 (0.03) -> 9.35 (0.13); 8.61 (0.27) -> 8.53 (0.15); 8.23 (0.23) -> 7.95 (0.05); 7.97 (0.22) -> 7.70 (0.06)Improved
Use ProxQuant-Ternary instead of TTQResNet ternary quantization · CIFAR-10top-1 classification error (%): 8.87 -> 8.40 (0.13); 7.63 -> 7.65 (0.15); 7.02 -> 7.05 (0.08)Mixed
Report the best of 4 ProxQuant-Ternary runs rather than the meanProxQuant-Ternary · CIFAR-10, ResNet ternarytop-1 classification error (%): 8.40 (0.13) -> 8.22; 7.65 (0.15) -> 7.53; 7.05 (0.08) -> 6.98Improved
Use ProxQuant-Binary instead of BinaryConnectBinary LSTM · Penn Treebankperplexity per word (PPW): 372.2 -> 288.5Improved
Use ALT-ProxQuant instead of ALT Straight-throughMulti-bit LSTM · Penn Treebank, 1-bitperplexity per word (PPW): 104.7 -> 106.2Hurt
Use ALT-ProxQuant instead of ALT Straight-throughMulti-bit LSTM · Penn Treebank, 2-bitperplexity per word (PPW): 90.2 -> 90.0Improved
Use ALT-ProxQuant instead of ALT Straight-throughMulti-bit LSTM · Penn Treebank, 3-bitperplexity per word (PPW): 86.1 -> 87.2Hurt
Scale the alternating quantizer output by 0.3 (ablation)ALT Straight-through LSTM · Penn Treebank, 1-, 2-, and 3-bitperplexity per word (PPW): 116.7, 94.3, 87.3 -> 104.7, 90.2, 86.1Improved

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