The current data management pages do not serve their purpose well. The Datasets page shows only a simple overview and the Annotations page, well, does not exist yet 😅
Since the annotations are already visible on the LLM eval / Crowdsourcing pages, I think we should focus more on datasets and model outputs.
In particular, it would be useful to have a brief overview for which datasets we have outputs from which models. Currently, the user can only click through individual examples on the Browse page, but there is no high-level view.
In the very least, we should hide the annotations management page from the front page.
The current data management pages do not serve their purpose well. The Datasets page shows only a simple overview and the Annotations page, well, does not exist yet 😅
Since the annotations are already visible on the LLM eval / Crowdsourcing pages, I think we should focus more on datasets and model outputs.
In particular, it would be useful to have a brief overview for which datasets we have outputs from which models. Currently, the user can only click through individual examples on the Browse page, but there is no high-level view.
In the very least, we should hide the annotations management page from the front page.