List your company

Question Answering on Musique (subEM, Original EM)

27subEMDSPO
9.8414.29518.7523.205Oct 10, 2025

Evaluation results

MethodLinks
DSPOModel Scale=7B, Dynamic Filter=true, Sequence-level Optimization=true2025.1027—
Search-R1 (GRPO, 14B)Model Scale=14B, RL Algorithm=GRPO2025.102619.1
DSPO w/o seq-lvl opt.Model Scale=7B, Dynamic Filter=true, Sequence-level Optimization=false2025.1013.3—
Search-R1 (GRPO, 7B)Model Scale=7B, RL Algorithm=GRPO2025.1012.216.2
DSPO w/o dyn. filterModel Scale=7B, Dynamic Filter=false, Sequence-level Optimization=true2025.1010.5—
Search-R1 (PPO, 14B)Model Scale=14B, RL Algorithm=PPO2025.10—21
Search-R1 (PPO, 7B)Model Scale=7B, RL Algorithm=PPO2025.10—14.6