Link the blog write-up at the top of the dataset card (#3)

- Link the blog write-up at the top of the dataset card (eedeeec783b10295cbc534b51bc231e20a43e210)
- Drop the Refiner link from the callout (971202253ddc43e2b98fd504cf435b057865d774)
- Add UTM tags to the macrodata.co links in Related Resources (396ddce78187a1d9c911a7bd2fe1814b6559296b)

README.md

@@ -27,6 +27,8 @@ configs:
 
 # WGO-Bench: What's Going On Benchmark
 
+> **Read the write-up:** [Segmenting Robot Video into Actionable Subtasks](https://macrodata.co/blog/annotating-robot-video-subtasks?utm_source=huggingface&utm_medium=dataset_card&utm_campaign=wgo-bench&utm_content=top_callout) covers how WGO-Bench was built, the 60+ annotation pipeline experiments we ran on it, and what worked.
+
 WGO-Bench is a small, manually annotated benchmark for evaluating how well vision-language models can turn robot and egocentric manipulation videos into timestamped subtask annotations.
 
 Each row contains one video episode, a high-level task instruction, and gold subtask segments with start time, end time, and a concise action label. The benchmark is designed for two related tasks:
@@ -118,6 +120,6 @@ The labels are short action descriptions grounded in what is visible in the clip
 
 ## Related Resources
 
-- Blog post: [Segmenting Robot Video into Actionable Subtasks](https://macrodata.co/blog/annotating-robot-video-subtasks)
-- Refiner example: [Subtask annotations](https://macrodata.co/docs/examples/annotations/subtask-annotations)
+- Blog post: [Segmenting Robot Video into Actionable Subtasks](https://macrodata.co/blog/annotating-robot-video-subtasks?utm_source=huggingface&utm_medium=dataset_card&utm_campaign=wgo-bench&utm_content=related_resources)
+- Refiner example: [Subtask annotations](https://macrodata.co/docs/examples/annotations/subtask-annotations?utm_source=huggingface&utm_medium=dataset_card&utm_campaign=wgo-bench&utm_content=related_resources)
 - Code: [macrodata-labs/refiner](https://github.com/macrodata-labs/refiner)