AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Further results on latent discourse models and word embeddings

Sammy Khalife, Douglas Gonçalves, Youssef Allouah, Leo Liberti

JMLR 2021 · Architectures and efficiency › RNNs and normalization · approach: Learned embeddings for retrieval · 1 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
check linear PMI vs scalar product relation (Eq. 25) with SN vectorsArora et al. latent discourse model · corpus 1, d=50, window 2Pearson correlation: 0.04 (R2 0.002)Hurt
increase dimension from 50 to 300 for Eq. 25 on corpus 1SN embeddings, window 2 · corpus 1Pearson correlation: 0.04 -> 0.24Mixed
increase dimension from 50 to 300 for Eq. 23SN embeddings, window 2 · corpus 1Pearson correlation: 0.73 -> 0.76Improved
increase dimension from 50 to 300 for Eq. 24 (full regression)SN embeddings, window 2 · corpus 1Pearson correlation: 0.84 -> 0.91Improved
restrict regression to the 50 most frequent words (partial) for Eq. 24 (ablation)SN embeddings, window 10 · corpus 1, d=50slope: 0.120 -> 0.049Improved
increase window size from 2 to 10 for Eq. 25SN embeddings, d=50 · corpus 1Pearson correlation: 0.04 -> 0.09Mixed
use GloVe instead of SN vectors for Eq. 25word embeddings, window 2, d=50 · corpus 2Pearson correlation: 0.21 -> 0.05Hurt
use larger corpus (corpus 3 vs corpus 1) for Eq. 25 with SNSN embeddings, d=50, window 2 · Wikipedia extractsPearson correlation: 0.04 -> 0.21Improved
restrict word-rank range to top 5006 for the PMI-scalar-product relation (ablation)SN embeddings · corpus 2, d=50, window 2fit quality of Eq. 25Improved
change concentration assumption from Gaussian to symmetric bounded vectorspartition function concentration · theoryconcentration of ZcNo 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