Skip to content

Bench submission: https://github.com/TIGER-AI-Lab/StructEval #11

Description

@reacher-z

Benchmark source URL
https://github.com/TIGER-AI-Lab/StructEval

Anything else? (who owns it, what the score means; optional)
StructEval is maintained by TIGER-AI-Lab and evaluates LLM structured-output generation and cross-format conversion across 18 text and renderable formats and 44 task types. It reports syntax validity, structural correctness, and visual fidelity. Paper: https://arxiv.org/abs/2505.20139 Project: https://structeval.github.io/ Dataset: https://huggingface.co/datasets/TIGER-Lab/StructEval


We can provide a structured results export or public leaderboard link if needed.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions