Add methodology and results to dataset card

README.md

@@ -1,31 +1,37 @@
 ---
 license: cc-by-nc-nd-4.0
 task_categories:
-- image-to-text
-- document-question-answering
+  - image-to-text
+  - document-question-answering
 language:
-- en
+  - en
 size_categories:
-- n<1K
+  - n<1K
 tags:
-- selection-detection
-- checkbox-detection
-- benchmark
-- document-ai
-- ocr
+  - selection-detection
+  - checkbox-detection
+  - benchmark
+  - document-ai
+  - selection-f1
+  - ocr
 configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*.parquet
+  - config_name: default
+    data_files:
+      - split: train
+        path: data/train-*.parquet
+  - config_name: results
+    data_files:
+      - split: train
+        path: results/train-*.parquet
 ---
+
 # PulseBench-Select
 
-A benchmark for selection detection in document images.
+**A benchmark for selected-option detection in document images.**
 
-PulseBench-Select contains 485 document images with cleaned ground-truth annotations for checkbox, radio-button, and marked-option selection detection. Each sample pairs an image with a public `ground_truth` JSON object containing page-level annotations and a flattened `selected_items` list.
+PulseBench-Select contains 485 cleaned document images with ground-truth annotations for checkboxes, radio buttons, and marked answer choices. Each sample pairs a document image with public ground truth for the visible options that are selected.
 
-* Scoring methodology (GitHub): `https://github.com/Pulse-Software-Corp/PulseBench-Select`
+- **Scoring methodology (GitHub):** `https://github.com/Pulse-Software-Corp/PulseBench-Select`
 
 ## Quick Start
 
@@ -33,30 +39,95 @@ PulseBench-Select contains 485 document images with cleaned ground-truth annotat
 from datasets import load_dataset
 import json
 
+# Load benchmark data: document images plus cleaned ground truth.
 ds = load_dataset("pulse-ai/PulseBench-Select")
 
 sample = ds["train"][0]
-sample["image"]                         # PIL image
+sample["sample_id"]                    # Public sample id
+sample["image"]                        # PIL image of the document page
 gt = json.loads(sample["ground_truth"])
-gt["selected_items"]                    # selected options for evaluation
+gt["selected_items"]                   # Selected options used for scoring
+
+# Load aggregate benchmark results.
+results = load_dataset("pulse-ai/PulseBench-Select", "results")
+
+row = results["train"][0]
+row["display_name"]                    # Provider display name
+row["selection_f1"]                    # Corpus-pooled Selection F1
 ```
 
 ## Dataset Overview
 
+| Split | Samples | Page Annotations | Selected Items | Selection Candidates |
+|-------|---------|------------------|----------------|----------------------|
+| train | 485 | 14,516 | 1,976 | 4,180 |
+
 The benchmark focuses on pages where systems must determine which visible options are selected. Ground-truth coordinates are normalized eight-point polygons in reading order: `[x0, y0, x1, y1, x2, y2, x3, y3]`.
 
+459 samples contain at least one selected item; 26 samples contain no selected items and are retained to measure false positives.
+
+## Scoring: Selection F1
+
+Selection F1 evaluates only the positive selected class.
+
+1. **Parse** ground truth and predictions into selected-item records with `sample_id`, `page`, `content`, `bbox`, and `selected`.
+2. **Match** each predicted selected item to the best unmatched ground-truth selected item on the same sample and page.
+3. **Filter matches** with content token overlap >= 0.80. If both items include 8-point bounding boxes, the bbox centroid distance must also be <= 0.35 in normalized page units.
+4. **Score** matched selected items as true positives, unmatched predictions as false positives, and unmatched ground-truth items as false negatives.
+5. **Report** corpus-pooled micro precision, recall, and F1, along with per-sample macro diagnostics.
+
+Token overlap is computed as:
+
+```text
+|tokens(ground_truth) intersect tokens(prediction)| / max(|tokens(ground_truth)|, |tokens(prediction)|)
+```
+
+The bbox centroid check is a veto used to prevent repeated labels with identical text, such as multiple `Yes` or `No` options on the same page, from matching the wrong spatial item. The public scorer supports disabling this check with `--centroid-max -1`.
+
+## Results
+
+We evaluated 6 systems using Selection F1. Scores below are corpus-pooled micro precision, recall, and F1 from the benchmark run associated with this release.
+
+| Rank | Provider | Precision | Recall | Selection F1 |
+|------|----------|-----------|--------|--------------|
+| 1 | **Pulse** | **0.782** | **0.761** | **0.772** |
+| 2 | GPT-5.5 | 0.383 | 0.311 | 0.343 |
+| 3 | Gemini 3.1 Pro | 0.334 | 0.317 | 0.325 |
+| 4 | Gemini 3.5 Flash | 0.317 | 0.311 | 0.314 |
+| 5 | Claude Opus 4.8 | 0.307 | 0.293 | 0.300 |
+| 6 | GPT-4o | 0.223 | 0.191 | 0.206 |
+
+The `results` config includes these aggregate results plus macro precision, macro recall, and skipped-sample counts for each provider.
+
 ## Schema
 
+### Default config
+
 | Column | Type | Description |
-|---|---|---|
+|--------|------|-------------|
 | `sample_id` | string | Stable public sample identifier |
 | `image` | image | Document image |
 | `ground_truth` | string | JSON with `page_count`, `annotations`, and `selected_items` |
 | `annotation_count` | int | Number of cleaned page annotations |
 | `selected_count` | int | Number of selected ground-truth items |
-| `selection_candidate_count` | int | Number of annotations/cells containing visible selection marks |
+| `selection_candidate_count` | int | Number of annotations or cells containing visible selection marks |
 | `selection_stats` | string | JSON summary for the row |
 
+### Results config
+
+| Column | Type | Description |
+|--------|------|-------------|
+| `rank` | int | Rank by corpus-pooled Selection F1 |
+| `provider` | string | Provider identifier |
+| `display_name` | string | Provider display name |
+| `precision` | float | Corpus-pooled positive-class precision |
+| `recall` | float | Corpus-pooled positive-class recall |
+| `selection_f1` | float | Corpus-pooled positive-class F1 |
+| `macro_precision` | float | Mean per-sample precision over scored samples |
+| `macro_recall` | float | Mean per-sample recall over scored samples |
+| `macro_skipped_samples` | int | Samples skipped from macro averaging because precision or recall was undefined |
+| `metric_version` | string | Metric version used for the reported row |
+
 ## Ground Truth Format
 
 ```json
@@ -83,10 +154,6 @@ The benchmark focuses on pages where systems must determine which visible option
 }
 ```
 
-## Scoring
-
-The public scorer computes precision, recall, and F1 over the positive selected class. A predicted selected item matches a ground-truth selected item when it is on the same sample/page and its content token overlap is at least 0.80.
-
 ## License
 
 This dataset is released under [CC BY-NC-ND 4.0](https://creativecommons.org/licenses/by-nc-nd/4.0/).

results/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:306be3b11d30fb24cacd15c30899bb0ad3f9b012adc88db863a1b037b830b64c
+size 3626