AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

Transformers learn factored representations

Adam Shai, Loren Amdahl-Culleton, Casper L. Christensen, Henry R Bigelow, Fernando Rosas, Alexander Boyd, Eric A. Alt, Kyle J Ray, Paul M. Riechers

ICML 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Equivariant and disentangled latent representations · 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
train on next-token prediction (vs initialization)GPT-2 style transformer · Five independent factorsdimensions for 95% variance: ~100 -> 11Improved
increase d_model to 480 (ablation)transformer · Five independent factorsdimensionalityNo clear effect
add noisy channel (epsilon>0) breaking conditional independence (ablation)transformer · Five factors with noisedimensionality over trainingMixed
use small epsilon (ablation)transformer · Noisy-channel factorsdimensionalityNo clear effect
use large epsilon (ablation)transformer · Noisy-channel factorsdimensionalityImproved
train for longer on independent factorstransformer · Vary-one factor subspacessum of per-factor dims vs union: sum falls to 14, union 12Improved
replace transformer with RNN/LSTM (ablation)RNN, LSTM · Five independent factorsCEV curve factored-nessMixed

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