List your company

Recommendation on Netflix high reward trajectories

106.2NDCG@10Exp-RSFT
-107.52-52.0353.4558.935Mar 10, 2026

Evaluation results

MethodLinks
Exp-RSFT2026.03106.269.7198.0749.5886.3117.27
RSFT2026.03000000
PPO2026.03-4.88-24.01-14.66-35.71-8.613.39
DPO2026.03-99.3-96.77-99.14-95.52-95.3419.21