VESPO: Variational Sequence-Level Soft Policy Optimization for Stable Off-Policy LLM Training
Paper
•
2602.10693
•
Published
•
4
floyed shen
floyed
AI & ML interests
None yet
Recent Activity
authored
a paper
1 day ago
Safety Instincts: LLMs Learn to Trust Their Internal Compass for Self-Defense
authored
a paper
1 day ago
VESPO: Variational Sequence-Level Soft Policy Optimization for Stable Off-Policy LLM Training
authored
a paper
1 day ago
StressPrompt: Does Stress Impact Large Language Models and Human
Performance Similarly?
Organizations
