atlas

Research map / Image generation / Text-to-image diffusion

Tractable Control for Autoregressive Language Generation

ICML 2023 · Image generation › Text-to-image diffusion · approach: Text generation and structured prediction · 2 citations (OpenAlex)

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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 GeLaTo with tractable HMM constraint guidance instead of A*esque decodingGPT2-large on CommonGen, unsupervised · CommonGen testBLEU-4: 28.6 -> 29.0Helped
Use GeLaTo with tractable HMM constraint guidance instead of A*esque decodingGPT2-large on CommonGen, supervised · CommonGen testBLEU-4: 28.2 -> 34.1Helped
Use GeLaTo rather than NADOYelp!Review · Yelp!ReviewBLEU-4: 6.0 -> 6.6Helped
Use GeLaTo rather than InsNetNews · NewsBLEU-4: 5.0 -> 5.4Helped
Increase beam size (ablation)GeLaTo on CommonGen · CommonGen devBLEUHelped
Vary the supervised mixture weight w across 0.1 to 0.8 (ablation)GeLaTo on CommonGen · CommonGen devBLEUNo clear effect
Improve the HMM approximation of the base language model through additional distillation training (ablation)GeLaTo · CommonGenBLEUHelped

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