Wang et al., HelpSteer2: Open-source dataset for training top-performing reward models (2024). NVIDIA and Scale AI.
Access
Human-rated evaluation responses. Source does not identify the individual NVIDIA model for each response; model identity is explicitly unspecified. Prompts are mostly ShareGPT-derived.
Privacy handling
Only prompt/response text is retained and privacy-scanned; model identities and human ratings are not invented.
Last reviewed
Nov 15, 2026
Imported records are labelled with their source, keep upstream attribution and license, and never earn Research Points.
I would like you to audit content for me based on a URL I specify below. I would like you to audit content as if you were a Google Quality Rater following the rules set out by Google (which you can see here…
1 turnen
Unspecified NVIDIA model (HelpSteer2)· nvidia NVIDIA HelpSteer2
I would like you to audit content for me based on a URL I specify below. I would like you to audit content as if you were a Google Quality Rater following the rules set out by Google (which you can see here…