Dataset · Hugging Face

songjhPKU/PM4Bench

Name: PM4Bench
Creator: songjhPKU
License: apache-2.0
Keywords: task_categories:visual-question-answering, task_categories:image-to-text, source_datasets:MMMU/MMMU_Pro, source_datasets:laolao77/MMDU, language:ar, language:cs, language:en, language:hu, language:ko, language:ru, language:sr, language:th, language:vi, language:zh, license:apache-2.0

PM4Bench Strictly parallel multilingual evaluation for Large Vision-Language Models Overview The paper Benchmarking and Boosting Multilingual Capabilities of LVLMs via OCR-Centric Reinforcement Learning introduces PM4Bench to separate language effects from dataset variation. Its content is strictly parallel across ten languages, and its vision setting renders textual inputs directly into images. Comparing that setting with interleaved input identifies OCR… See the full description on the dataset page: https://huggingface.co/datasets/songjhPKU/PM4Bench.

View on Hugging FaceReport an issue

Author / organization	songjhPKU
License (as declared)	apache-2.0
Downloads	3K (as reported by Hugging Face)
Likes	2
Created	18 February 2025
Last updated at source	25 August 2026

multilingualvision-languageocrgui-grounding

Embed this entry

<iframe src="https://silytics.si/embed/item/c9287f7b-c8f5-46b9-a3e8-56ea59d3f9fe" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

3,015 now (-319 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from songjhPKU

songjhPKU/ScannedRxnDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.