SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning
Paper
•
2506.19767
•
Published
•
12
SONGJUN TU
SONGJUNTU
·
AI & ML interests
None yet
Recent Activity
authored
a paper
2 days ago
In-Dataset Trajectory Return Regularization for Offline Preference-based
Reinforcement Learning
authored
a paper
2 days ago
Enhancing LLM Reasoning with Iterative DPO: A Comprehensive Empirical
Investigation
authored
a paper
2 days ago
SRFT: A Single-Stage Method with Supervised and Reinforcement
Fine-Tuning for Reasoning
Organizations
