📊

Dataset

EditCLEVR

Name: EditCLEVR
Creator: torux
License: CC-BY-4.0

by torux torux/editclevr

Free2AITools Nexus Index

60.8

S: Semantic 50

Query-time baseline · scored live at search

A: Authority 62

P: Popularity 54

R: Recency 86

Q: Quality 50

Tech Context

Vital Performance —

Source →

Data Integrity 60.8 FNI Score

- Size

- Rows

- Tokens

Dataset Information Summary
Entity Passport
Registry ID	torux/editclevr
License	CC-BY-4.0
Provider	huggingface

📜

Cite this dataset

Academic & Research Attribution

BibTeX

@misc{hf_dataset_torux_editclevr,
  author = {torux},
  title = {EditCLEVR Dataset},
  year = {2026},
  howpublished = {\url{https://huggingface.co/datasets/torux/EditCLEVR}},
  note = {Accessed via Free2AITools.}
}

APA Style

torux. (2026). EditCLEVR [Dataset]. Free2AITools. https://huggingface.co/datasets/torux/EditCLEVR

🔬Technical Deep Dive

Full Specifications [+]

⚖️ Free2AITools Nexus Index V2.0

Methodology How FNI works

Semantic (S) 50

Query-time baseline · scored live at search

Authority (A) 62

Popularity (P) 54

Recency (R) 86

Quality (Q) 50

💬 Index Insight

FNI V2.0 for EditCLEVR: Authority (A:62), Popularity (P:54), Recency (R:86), Quality (Q:50). Semantic (S) is a query-time baseline scored live at search.

Free2AITools Nexus Index

Data Sources / Provenance

HuggingFace API GitHub Metadata Arxiv Citation DB Methodology

Open data Updated: Live data

⬇️

Downloads

50,139

🎯 Task Categories

image-to-image

👁️ Data Preview

📊

Row-level preview not available for this dataset.

Schema structure is shown in the Field Logic panel when available.

🔗 Explore Full Dataset ↗

🧬 Field Logic

🧬

Schema not yet indexed for this dataset.

Dataset Specification

EditCLEVR Phase 1

EditCLEVR is a synthetic benchmark for evaluating intervention faithfulness in object-centric representations. Each example is a paired before/after scene where exactly one object-level factor may change.

Dataset summary

~20k paired edits across six evaluation splits
Suites: atomic single-factor edits, no-edit controls, hard distractors, and CoGenT-OOD combinations
Per-object factors: color, material, size, shape
Artifacts per pair: before/after RGB images, instance masks (.npz), scene JSON, object attributes, edit metadata, and difficulty tags

Splits

Split	Purpose
`train`	Probe training
`val`	Validation
`test_id`	In-distribution atomic edits
`test_noop`	No-edit control pairs
`test_hard`	Hard distractor edits
`test_cogent`	CoGenT-OOD edits

Download

The dataset ships as a small set of .tar.gz archives (one per suite plus a splits bundle). The helpers below download and extract them into the original directory layout automatically.

bash

git clone https://github.com/torux-bughunter/EditCLEVR.git
cd EditCLEVR
pip install -e ".[hub]"
python -m editclevr.download

Or from Python:

python

from editclevr.download import setup_dataset

setup_dataset()

Evaluate

bash

pip install -e .
python3 -m editclevr.evaluation.run_evaluation

File layout

text

splits.json
atomic_id/
no_edit/
hard_distractor/
cogent_ood/
validation_report.json
phase1_manifest.json

splits.json stores relative paths to images, masks, and scene JSON files so the dataset can be moved across machines.

License

Dataset: CC-BY-4.0
Code: [MIT](https://github.com/torux-bughunter/

Social Proof

HuggingFace Hub

50.1KDownloads

Hub Discussions

🤗 Data Source: Hugging Face ↗

🔄 Updated daily

Source summary: Based on Hugging Face metadata. Not a recommendation.

📊 FNI Methodology 📚 Knowledge Baseℹ️ Verify with original source

🛡️ Dataset Transparency Report

Technical metadata sourced from upstream repositories.

Open Metadata

🆔 Identity & Source

id: hf-dataset--torux--editclevr
slug: torux--editclevr
source: huggingface
author: torux
license: CC-BY-4.0
tags: task_categories:image-to-image, language:en, license:cc-by-4.0, size_categories:10k<n<100k, region:us, object-centric-learning, intervention-faithfulness, clevr, synthetic, computer-vision

⚙️ Technical Specs

architecture: null
params billions: null
context length: null
pipeline tag

📊 Engagement & Metrics

downloads: 50,139
stars: null
forks: null

Data indexed from public sources. Updated daily.