AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Trellis Networks for Sequence Modeling

Shaojie Bai, J. Zico Kolter, Vladlen Koltun

ICLR 2019 · Architectures and efficiency › RNNs and normalization · approach: RNN training dynamics and gradients · 94 citations (OpenAlex)

This paper tried a common change: Add a reconstruction / auxiliary loss

add auxiliary losses · Word-level PTB test · perplexity: 57.99 -> 56.97 Improved

What 987 accepted papers report for this change

See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.

Read the paper

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 TrellisNet-MoS (1.4x larger)Word-level PTB · Word-level PTB testperplexity: 56.97 -> 54.19Improved
use TrellisNet instead of Relational Memory CoreWord-level WT103 · Word-level WT103 testperplexity: 31.6 -> 29.19Improved
use TrellisNet instead of AWD-LSTMCharacter-level PTB · Character-level PTB testbits-per-character: 1.175 -> 1.158Improved
use TrellisNet instead of IndRNNSequential MNIST · Sequential MNISTaccuracy: 99.0 -> 99.20Improved
use TrellisNet instead of TransformerPermuted MNIST · Permuted MNISTaccuracy: 97.9 -> 98.13Improved
use TrellisNet instead of r-LSTM with auxiliary lossSequential CIFAR-10 · Sequential CIFAR-10accuracy: 72.2 -> 73.42Improved
add hidden (VD-based) dropout (ablation)TrellisNet · Word-level PTB testperplexity: 64.69 -> 56.97Improved
add weight dropout (ablation)TrellisNet · Word-level PTB testperplexity: 63.82 -> 56.97Improved
add auxiliary losses (ablation)TrellisNet · Word-level PTB testperplexity: 57.99 -> 56.97Improved
add long-sequence parallelism (ablation)TrellisNet · Word-level PTB testperplexity: 57.35 -> 56.97Improved
use a dense kernel instead of mixed group convolution (ablation)TrellisNet · Word-level PTB testperplexity: 59.18 -> 56.97Improved
inject input at every layer rather than every 2 layers (ablation)TrellisNet · Word-level PTB testperplexity: 57.44 -> 56.97Improved

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