LiQiiiii/DLLM-Survey

A survey's companion shelf, with the paper's title and a quoted citation card up top, two wide figures from the paper, then a long table of contents and thirty-three narrow two-column lists of papers, each with a way back to the top.

github.com/LiQiiiii/DLLM-SurveyLiQiiiii/DLLM-Survey
Screenshot of https://github.com/LiQiiiii/DLLM-Survey, © the authors. Shown for reference. Request removal
Venue
TMLR 2026
Year
2026
First author
Runpeng Yu
Field
discrete diffusion language and multimodal models
Template
Centred h2 paper title and h4 subtitle, a lone teal arXiv badge over a row of four status badges (stars, forks, last commit, PRs Welcome), a left-aligned blockquote citation card with authors, venue and affiliation, two full-width paper figures with bold numbered captions
Theme
Light

Checked against the rendered page. The taxonomy figure points at assets/taxonomy.png, which is not in the repository, so it renders as an empty gap. The stars and forks badges still use the old repository name and only render through GitHub's redirect. One Vision and Multimodal row repeats the title of the row above with different authors, and Quick Links leaves out Reinforcement Learning.

Why it is here

Added on the maintainer's suggestion rather than through the design rounds, since the survey has no project page for the README pool to follow. It reads like a paper's companion page, with the title and a citation card in a blockquote, two large figures from the paper, and then 128 papers laid out in 33 identical two-column tables behind a long table of contents.

Overview

Header Block

Badge Colors

Section Outline

Guidelines

More like this