"OneLA: Scaling Linear-Attention Decoding to Large Beams in Generative Recommendation" scored 6.0/10 predicted impact on @kurateorg — Clarity 7.5 · Difficulty 7.0 · Translational 7.0