Hugging Face's logo

Datasets:
top-papers
/
top-papers-graph-benchmark

Modalities:
Image
Formats:
imagefolder
Languages:
Russian
Size:
1K - 10K
Tags:
scientific-reasoning
multimodal
vision-language
benchmark
russian
Libraries:
Datasets
Dataset card Data Studio Files
xet
 Community
1
top-papers commited on
Commit
8937fed
·
verified ·
1 Parent(s): 8da87fc

Upload Task 3 VLM benchmark dataset from Colab

Browse files

Generated from Task 3 Google Sheet URLs and adapted for VLM A/B generation.

Files changed (4) hide show
  1. data/task3_cases_flat.jsonl +0 -0
  2. data/task3_vlm_generation.jsonl +0 -0
  3. metadata/build_summary.json +2 -2
  4. review_metadata/task3_case_rationales.jsonl +0 -0
data/task3_cases_flat.jsonl CHANGED
The diff for this file is too large to render. See raw diff
 
data/task3_vlm_generation.jsonl CHANGED
The diff for this file is too large to render. See raw diff
 
metadata/build_summary.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "generated_at": "2026-05-12T18:22:14Z",
3
  "dataset_repo_id": "top-papers/top-papers-graph-benchmark",
4
  "dataset_config_name": "task3_vlm_generation",
5
  "split_name": "test",
@@ -30,7 +30,7 @@
30
  ],
31
  "explicit_images": [],
32
  "explicit_pdfs": [],
33
- "other_files_discovered": 12,
34
  "manifests_discovered": 20,
35
  "cases_total": 210,
36
  "cases_written": 182,
 
1
  {
2
+ "generated_at": "2026-05-16T09:07:29Z",
3
  "dataset_repo_id": "top-papers/top-papers-graph-benchmark",
4
  "dataset_config_name": "task3_vlm_generation",
5
  "split_name": "test",
 
30
  ],
31
  "explicit_images": [],
32
  "explicit_pdfs": [],
33
+ "other_files_discovered": 23,
34
  "manifests_discovered": 20,
35
  "cases_total": 210,
36
  "cases_written": 182,
review_metadata/task3_case_rationales.jsonl CHANGED
The diff for this file is too large to render. See raw diff