README.md

---
dataset_info:
- config_name: Empathetic_Captioning
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: social_attribute
    dtype: string
  - name: simple_prompt
    dtype: string
  - name: empathic_prompt
    dtype: string
  splits:
  - name: train
    num_bytes: 5064542
    num_examples: 204
  download_size: 4980082
  dataset_size: 5064542
- config_name: Image_Resilience
  features:
  - name: ID
    dtype: string
  - name: attack_type
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: Question
    dtype: string
  - name: Answer
    dtype: string
  splits:
  - name: train
    num_bytes: 13954350
    num_examples: 1500
  download_size: 12590814
  dataset_size: 13954350
- config_name: Instance_Identity
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: Question
    dtype: string
  - name: Answer
    dtype: string
  splits:
  - name: train
    num_bytes: 12066443
    num_examples: 1343
  download_size: 10907623
  dataset_size: 12066443
- config_name: Multilingual_CloseEnded
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: Question
    dtype: string
  - name: Options
    dtype: string
  - name: Answer
    dtype: string
  - name: Reasoning
    dtype: string
  splits:
  - name: train
    num_bytes: 51228680
    num_examples: 625
  download_size: 50857340
  dataset_size: 51228680
- config_name: Multilingual_OpenEnded
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: Question
    dtype: string
  - name: Answer
    dtype: string
  splits:
  - name: train
    num_bytes: 5507897
    num_examples: 625
  download_size: 5000097
  dataset_size: 5507897
- config_name: Multiple_Choice_VQA
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: Question
    dtype: string
  - name: Options
    dtype: string
  - name: Answer
    dtype: string
  - name: Reasoning
    dtype: string
  splits:
  - name: train
    num_bytes: 128839429
    num_examples: 1844
  download_size: 127658673
  dataset_size: 128839429
- config_name: Scene_Understanding
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: Attribute
    dtype: string
  - name: version_type
    dtype: string
  - name: Question
    dtype: string
  - name: Answer
    dtype: string
  splits:
  - name: train
    num_bytes: 122328629
    num_examples: 13668
  download_size: 109845740
  dataset_size: 122328629
- config_name: Visual_Grounding
  features:
  - name: ID
    dtype: string
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: bbox
    list: float64
  splits:
  - name: train
    num_bytes: 33346641
    num_examples: 285
  download_size: 33310075
  dataset_size: 33346641
configs:
- config_name: Empathetic_Captioning
  data_files:
  - split: train
    path: Empathetic_Captioning/train-*
- config_name: Image_Resilience
  data_files:
  - split: train
    path: Image_Resilience/train-*
- config_name: Instance_Identity
  data_files:
  - split: train
    path: Instance_Identity/train-*
- config_name: Multilingual_CloseEnded
  data_files:
  - split: train
    path: Multilingual_CloseEnded/train-*
- config_name: Multilingual_OpenEnded
  data_files:
  - split: train
    path: Multilingual_OpenEnded/train-*
- config_name: Multiple_Choice_VQA
  data_files:
  - split: train
    path: Multiple_Choice_VQA/train-*
- config_name: Scene_Understanding
  data_files:
  - split: train
    path: Scene_Understanding/train-*
- config_name: Visual_Grounding
  data_files:
  - split: train
    path: Visual_Grounding/train-*
---



# Humanibench_Greek

This dataset is a Greek translation of [HumaniBench](https://huggingface.co/datasets/vector-institute/HumaniBench), a human-centered benchmark for evaluating Large Multimodal Models.

The goal of this version is to make HumaniBench usable as an evaluation benchmark for Greek-capable Vision-Language Models.

## Dataset Description

The dataset keeps the same task structure as the original HumaniBench dataset. The text fields have been translated into Greek.

HumaniBench contains multiple multimodal evaluation tasks, including visual question answering, multiple-choice VQA, multilingual QA, visual grounding and empathetic captioning.

## Available Tasks

The translated dataset follows the same task subsets as the original dataset:

* `Scene_Understanding`
* `Instance_Identity`
* `Multiple_Choice_VQA`
* `Multilingual_OpenEnded`
* `Multilingual_CloseEnded`
* `Visual_Grounding`
* `Empathetic_Captioning`
* `Image_Resilience`

## Intended Use

This dataset is intended for evaluating Vision-Language Models on human-centered multimodal tasks in Greek.

It can be used to test whether a model can understand images, answer questions, reason about visual content, handle multiple-choice prompts, and remain robust across different human-centered evaluation settings.

## Citation

Please cite the original HumaniBench paper when using this dataset:

```bibtex
@article{raza2025humanibench,
  title={HumaniBench: A Human-Centric Framework for Large Multimodal Models Evaluation},
  author={Raza, Shaina and Narayanan, Aravind and Khazaie, Vahid Reza and Vayani, Ashmal and Chettiar, Mukund S. and Singh, Amandeep and Shah, Mubarak and Pandya, Deval},
  journal={arXiv preprint arXiv:2505.11454},
  year={2025}
}
```