AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Compositional Generalization from First Principles

Thaddäus Wiedemer, Prasanna Mayilvahanan, Matthias Bethge, Wieland Brendel

NeurIPS 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Contrastive representation theory · 4 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 a compositional model instead of a monolithic modelMulti-sprite reconstruction with orthogonal training support · All latent spaceMSE: 4.06e-02 -> 1.24e-03Improved
Use a compositional model instead of a monolithic modelMulti-sprite reconstruction with random training · All latent spaceMSE: 1.73e-3 -> 1.07e-03Improved
Sample training from a normal distribution with orthogonal support (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.24e-03 -> 1.24e-03No clear effect
Use a broad diagonal training support (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.24e-03 -> 1.39e-03No clear effect
Broaden the orthogonal training support (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.24e-03 -> 1.16e-03No clear effect
Broaden the diagonal training support (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.39e-03 -> 1.13e-03No clear effect
Introduce gaps in the training support (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.24e-03 -> 2.64e-03Hurt
Narrow the diagonal training support so some background pixels remain occluded (ablation)Compositional model · Multi-sprite, all latent spaceMSE: 1.39e-03 -> 1.04e-02Hurt
Use RGBa component outputs and alpha compositing (ablation)Compositional model · Multi-sprite, orthogonal support; all latent spaceMSE: 1.24e-03 -> 5.24e-04Improved

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