Datasets:
Are flattened tables (single column) and stripped reference lists intentional?
#12
by SandyAndiowe - opened
Thanks for the amazing work on this dataset—it's a fantastic resource!
While working with the data recently, I noticed a couple of formatting behaviors and wanted to check if they are expected:
Table Formatting: It appears that in Nemotron CC & Nemotron CC v2, tables are being processed and flattened into a single vertical text column.
Missing References: The reference lists/citations in many of the data rows seem to be completely cleared out.
Could anyone confirm if these are intentional design choices for the data pipeline?
Thanks！