Rosinality(@rosinality):https://t.co/NxJ1EM1qhY Improving exploration for rubric RL using OPSD to generate guidance for underexplored or penalized criteria. As combining criteria to construct rewards is common maybe it could be useful beyond rubric RL.

2026.07.21 10:56

Improving exploration for rubric RL using OPSD to generate guidance for underexplored or penalized criteria. As combining criteria to construct rewards is common maybe it could be useful beyond rubric RL.

106

轉發到社區

熱門用戶

2M 粉絲

260.1K 粉絲

101.7K 粉絲

579K 粉絲

40.1K 粉絲

4.3M 粉絲

26.4M 粉絲

28.3K 粉絲

182.9K 粉絲

549.7K 粉絲

3.4M 粉絲

140.1K 粉絲

111K 粉絲

2.1M 粉絲

233.1K 粉絲