PoLi-RL: Point-to-List RL Boosts Conditional Semantic Similarity
PoLi‑RL, a point‑to‑list RL framework, achieved a Spearman score of 48.18 on the C‑STS benchmark by using a two‑stage curriculum that starts with pointwise rewards and adds hybrid rewards. Read more: getnews.me/poli-rl-point-to-list-rl... #polirl #csts
0
0
0
0