trapo-semi-supervised-rl

【数据与研究】Bridge labeled and unlabeled data through trajectory similarity in reinforcement learning. Select reliable unlabeled samples by comparing pass-rate evolution trajectories against labeled data. Achieve 42.6% accuracy with 1K labeled + 3K unlabeled samples, surpassing fully-supervised training on 45K labels with 10% annotation budget.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准