Efficient Self-Evaluation for Diffusion Language Models via Sequence Regeneration

A tidy stack of rounded cards washed in a blue-violet-apricot dawn, signed with a gradient wordmark.

zhongzero.github.io/DiSEEfficient Self-Evaluation for Diffusion Language Models via Sequence Regeneration
Screenshot of https://zhongzero.github.io/DiSE/, © the authors. Shown for reference. Request removal
Venue
ACL 2026
Year
2026
First author
Linhao Zhong
Field
diffusion language models
Template
Bootstrap 5.3 with custom style.css (template originally by Ximing Xing)
Theme
Light

Checked against the rendered page. Playfair Display is declared for the title and wordmark but not loaded; both render in Georgia.

Why it is here

Gradient wordmark logo and serif title, with content in soft-tinted rounded cards and a display-math block. A coherent card-based style that adds variety.

Overview

Color Palette

Typography

Avenir Next Cyr
Efficient Self-Evaluation for Diffusion Language Models via Sequence R Abstract · Method · Results · BibTeX — 0123456789
Georgia
Efficient Self-Evaluation for Diffusion Language Models via Sequence R Abstract · Method · Results · BibTeX — 0123456789
KaTeX fonts
Efficient Self-Evaluation for Diffusion Language Models via Sequence R Abstract · Method · Results · BibTeX — 0123456789

Spacing & Shape

Page Modules

Guidelines

More like this