α
Research
Alpha Leak
Conferences
Topics
Top Authors
Rankings
Browse All
EN
中
Home
/
Authors
/
Songjun Tu
Songjun Tu
2
papers
4
total citations
papers (2)
In-Dataset Trajectory Return Regularization for Offline Preference-based Reinforcement Learning
AAAI 2025
arXiv
3
citations
AlphaDecay: Module-wise Weight Decay for Heavy-Tailed Balancing in LLMs
NEURIPS 2025
arXiv
1
citations