VESPO: Variational Sequence-Level Soft Policy Optimization for Stable Off-Policy LLM Training
Paper • 2602.10693 • Published • 220
Julian Kindel
JulianKindel
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 2 months ago
ProAct: Agentic Lookahead in Interactive Environments upvoted a paper 2 months ago
RLAnything: Forge Environment, Policy, and Reward Model in Completely Dynamic RL SystemOrganizations
upvoted a paper 23 days ago
upvoted 2
papers 2 months ago
upvoted an article 3 months ago
Article
Scaling OpenEnv: From Free Usage to Thousands of Concurrent Environments
•
12
updated a
model 7 months ago
published 2
models 7 months ago