mit-han-lab/SMEPO download history
mit-han-lab/SMEPO is a text generation dataset on the Hugging Face Hub. In the last 30 days it was downloaded 74 times (30 in the last 7 days), and 735 times in total. It ranks #190,940 among datasets by monthly downloads.
This repository contains the datasets for the paper Hide to Guide: Learning via Semantic Masking. Official GitHub Repository: mit-han-lab/SMEPO Overview Semantic Masked Expert Policy Optimization (SMEPO) is an expert-guided Reinforcement Learning with Verifiable Rewards (RLVR) method.
Open mit-han-lab/SMEPO on Hugging Face
Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.