tlc4418/gold_labelled_gens

Dataset of 12600 answer generations from a 1.4b fine-tuned Pythia policy model, using the AlpacaFarm dataset 'val' split, and labelled with the AlpacaFarm 'reward-model-human' to give 'gold' scores. Used during best-of-n inference in 'Reward Model Ensembles Mitigate Overoptimization'

<iframe src="https://silytics.si/embed/item/a2cffba6-493e-4e60-99db-a78806ac4710" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.