Datasets:

davanstrien
/

dataset-paper-pages-smoke-extracted

arxiv_id stringclasses 8 values	page int32 1 29	n_pages int32 20 72	extraction stringlengths 1.7k 18.1k	inference_info stringclasses 1 value
2510.12565	3	30	{"focal_dataset_name": "MMOT (Ours)", "claimed_firsts": ["MMOT is the first large-scale drone-based multispectral MOT dataset"], "caption": "Table 1: Comparison of representative datasets. †: The statistics are computed based on publicly available labels.", "dimensions": ["Scenario", "Videos", "Total Frames", "Total Du...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2510.12565	5	30	{"focal_dataset_name": "MMOT (Ours)", "claimed_firsts": [], "caption": "Table 2: Density and motion statistics across datasets. Max: maximum objects per frame; @300px: average number of objects within a 300-pixel radius. Displacement is decomposed into drone (platform motion), object (ground-relative motion after compe...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2510.12565	26	30	{"focal_dataset_name": null, "claimed_firsts": [], "caption": "Table 8: Comparison of Input/Output Shapes and Parameters among Different Stem Designs", "dimensions": ["Component", "Original Stem", "2D-Stem (MSI)", "Spectral 3D-Stem (MSI)"], "rows": [{"dataset_name": "Input Shape", "is_focal": "no", "year": null, "value...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2510.12565	29	30	{"focal_dataset_name": null, "claimed_firsts": [], "caption": "Table 10: Comparison of tracking-by-detection methods under two different detectors (YOLOv11-L and Deformable-DETR) on the MMOT dataset. Metrics are reported for class-averaged and detection-averaged settings.", "dimensions": ["HOTA", "MOTA", "IDF1", "DetA"...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2511.17441	5	22	{"focal_dataset_name": "RoboCOIN", "claimed_firsts": [], "caption": "TABLE II: Qualitative comparison of existing real-world datasets for robot manipulation. All data is drawn from the original paper or the RoboMIND paper. N/A indicates that the dataset does not provide explicit annotations for the corresponding attrib...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2605.31086	2	32	{"focal_dataset_name": "RHELM (Ours)", "claimed_firsts": [], "caption": "Table 1: Comparison of representative AI memory benchmarks evaluated across multiple dimensions. Statistics are either from the original paper or based on our estimations.", "dimensions": ["Avg. Turns", "Context Window", "External Data", "Profile ...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2506.07671	1	20	{"focal_dataset_name": "GaRAGE", "claimed_firsts": [], "caption": "Table 1: Features of the GaRAGE benchmark relative to existing RAG benchmarks.", "dimensions": ["Human Intervention", "Question", "Answer", "Grounding"], "rows": [{"dataset_name": "MultiHop RAG (Tang and Yang, 2024)", "is_focal": "no", "year": "2024", "...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2602.17003	1	28	{"focal_dataset_name": "PERSONA2WEB", "claimed_firsts": ["the first personalized web agent benchmark"], "caption": "Table 1. Comparison between PERSONA2WEB and existing web agent benchmarks. PERSONA2WEB offers extensive domain and website coverage, enables evaluation on the open web, and supports personalization assess...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2605.02003	1	72	{"focal_dataset_name": "RamanBench", "claimed_firsts": [], "caption": "Table 1: Existing benchmarks and dataset collections for ML on Raman spectroscopy. Red cells mark a violation of RamanBench's inclusion criteria (synthetic or non-Raman data, restricted/unavailable access); yellow signals a partial concern (multi-sp...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2512.23437	6	34	{"focal_dataset_name": "RealX3D", "claimed_firsts": [], "caption": "Table 1: Comparison of existing degraded 3D datasets. RealX3D surpasses existing datasets in diversity and resolution, and further offers RAW sensor data that preserves richer signals under severe degradations, alongside high-end laser scans for precis...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]
2603.08147	1	31	{"focal_dataset_name": "MV-Fashion", "claimed_firsts": [], "caption": "Table 1. We compare the proposed MV-Fashion dataset with existing datasets, which we divide into two categories: (a) multi-view human, and (b) fashion-centric datasets. For the multi-view human comparison, we evaluate attributes such as Size (inclus...	[{"model_id": "numind/NuExtract3", "model_name": "NuExtract3", "column_name": "extraction", "timestamp": "2026-06-01T10:31:53.244958", "mode": "structured-extraction", "has_template": true, "enable_thinking": false, "temperature": 0.2, "max_tokens": 8192, "parse_error_rate": 0.0}]

NuExtract3 on davanstrien/dataset-paper-pages-smoke-comparison

This dataset contains outputs from davanstrien/dataset-paper-pages-smoke-comparison processed with NuExtract3, a 4B vision-language model for document understanding.

Processing Details

Source Dataset: davanstrien/dataset-paper-pages-smoke-comparison
Model: numind/NuExtract3
Mode: structured-extraction
Number of Samples: 11
Processing Time: 5.2 min
Processing Date: 2026-06-01 10:31 UTC

Configuration

Image Column: image
Output Column: extraction
Dataset Split: train
Temperature: 0.2
Thinking Mode: disabled

Extraction Template

{
  "focal_dataset_name": "verbatim-string",
  "claimed_firsts": [
    "verbatim-string"
  ],
  "caption": "verbatim-string",
  "dimensions": [
    "verbatim-string"
  ],
  "rows": [
    {
      "dataset_name": "verbatim-string",
      "is_focal": [
        "yes",
        "no"
      ],
      "year": "verbatim-string",
      "values": [
        {
          "dimension": "verbatim-string",
          "value": "verbatim-string"
        }
      ]
    }
  ]
}

Dataset Structure

Original columns plus:

extraction: NuExtract3 output (JSON string)
inference_info: JSON list tracking models applied to this dataset

Generated with UV Scripts

Downloads last month: 27

Number of rows:

Total file size:

4.45 MB