Skip to page content
SAVRN
Search Plan a project

Dataset · Visual question answering

PixelProof-Evaluation

by Project PixelProof PixelProof/PixelProof-Evaluation

The frozen evaluation sets for the PixelProof paper. Each eligible replay-verified world contributes exactly five deterministic, pixel-distinct instances.

Rows21,720
Configurations3
Size161.2 MB
License—
AccessPublicly accessible
Monthly Downloads6

Dataset Card

The frozen evaluation sets for the PixelProof paper. Each eligible replay-verified world contributes exactly five deterministic, pixel-distinct instances. The five instances of a world share its selected question text while varying the rendered scene and the gold answer, which the world's inverse program recovers from the image alone. The config names, the section column, and the ledgers/section files are holds spatially-steered worlds. The exclusions are the worlds that passed the 200-scene replay (1,301 profile-steered and 875 spatially-steered) but produced fewer than five pixel-distinct scenes; each one is documented in the corresponding ledgers/sectionexclusions.csv file. Each row…

Excerpt from the card by Project PixelProof.

Structure

combined_eval5 10,860 rows

SplitRowsSize
test10,860149.7 MB
evaluation_idstringsubset_versionstringselection_seedint64instance_rankint8sectionint8imageImageimage_sha256stringimage_widthint32image_heightint32questionstringanswerstringanswer_optionslistmarginfloat64quarantinedboolmodel_keystringgenerator_modelstringgenerator_harnessstringreasoning_effortstringprofilestringcampaign_idstringcandidate_idstringrecord_idstringexample_idstringprompt_familystringtask_titlestringreplay_statusstringmechanically_eligibleboolhuman_admissionstringcontainer_imagestringgeneration_seedint64source_image_pathstringsource_vlm_manifest_pathstringpixel_diversity_policystringpixel_support_scalestringpixel_support_shapestringpixel_diversity_cellstringpixel_attestation_statusstring

section1_eval5 6,490 rows

SplitRowsSize
test6,490109.6 MB
evaluation_idstringsubset_versionstringselection_seedint64instance_rankint8sectionint8imageImageimage_sha256stringimage_widthint32image_heightint32questionstringanswerstringanswer_optionslistmarginfloat64quarantinedboolmodel_keystringgenerator_modelstringgenerator_harnessstringreasoning_effortstringprofilestringcampaign_idstringcandidate_idstringrecord_idstringexample_idstringprompt_familystringtask_titlestringreplay_statusstringmechanically_eligibleboolhuman_admissionstringcontainer_imagestringgeneration_seedint64source_image_pathstringsource_vlm_manifest_pathstringpixel_diversity_policystringpixel_support_scalestringpixel_support_shapestringpixel_diversity_cellstringpixel_attestation_statusstring

section2_eval5 4,370 rows

SplitRowsSize
test4,37040.1 MB
evaluation_idstringsubset_versionstringselection_seedint64instance_rankint8sectionint8imageImageimage_sha256stringimage_widthint32image_heightint32questionstringanswerstringanswer_optionslistmarginfloat64quarantinedboolmodel_keystringgenerator_modelstringgenerator_harnessstringreasoning_effortstringprofilestringcampaign_idstringcandidate_idstringrecord_idstringexample_idstringprompt_familystringtask_titlestringreplay_statusstringmechanically_eligibleboolhuman_admissionstringcontainer_imagestringgeneration_seedint64source_image_pathstringsource_vlm_manifest_pathstringpixel_diversity_policystringpixel_support_scalestringpixel_support_shapestringpixel_diversity_cellstringpixel_attestation_statusstring

Details

Repository
PixelProof/PixelProof-Evaluation
Publisher
Project PixelProof
Task category
Visual question answering
Tags
synthetic, visual-question-answering, evaluation
Size category
Not stated by the source
Languages
Not stated by the source
Revision
b5d8ea04c365528c7599d091ac64ad1319a4e749
Last updated
2026-09-26

Files

14 files, 161.2 MB in total.

Data11 files · 161.2 MB
Documentation1 file · 3.1 KB
Other1 file · 1.2 KB
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
ledgers/section1_exclusions.csvData1.0 KB—
ledgers/section1_gold.csvData2.4 MB—
ledgers/section1_inputs.csvData6.8 MB—
ledgers/section2_exclusions.csvData394 B—
ledgers/section2_gold.csvData1.6 MB—
ledgers/section2_inputs.csvData4.6 MB—
manifests/mirror-validation.jsonData720 B—
manifests/release.jsonData796 B—
manifests/source-sections.jsonData2.4 KB—
parquet/section1/test-00000.parquetData107.3 MBfe7b1eccf36c
parquet/section2/test-00000.parquetData38.6 MBedb6bd7bfa81
README.mdDocumentation3.1 KB—
CHECKSUMS.sha256Other1.2 KB—
.gitattributesRepository2.5 KB—

License and Download

License
Not stated by the source
Access
No access gate
Download from Project PixelProof

Released by Project PixelProof through its official repository on Hugging Face.