Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Variational Proximal Policy Optimization

About

Reinforcement Learning from Human Feedback via Proximal Policy Optimization often suffers from policy mode collapse, brittle exploration loops, and distribution drift. This paper introduces Variational Proximal Policy Optimization (\(\textsc{VP}_2\textsc{O}\)), a particle-based variational inference framework that maps policy optimization to Stein Variational Gradient Descent within a Mixture-of-Experts architecture. By leveraging functional kernels over localized expert prototypes alongside an expert orthogonalization loss, \(\textsc{VP}_2\textsc{O}\) introduces a geometry-based proximal-control mechanism that can reduce reliance on fixed clipping or KL schedules. Our results on a 33B/4B sparse Mixture-of-Experts model show several improvements across complex reasoning benchmarks, establishing a \(+\mathbf{179}\) ELO gain on Codeforces and a \(\mathbf{32\%}\) reduction in token count on AIME mathematical reasoning tasks.

Ousmane Amadou Dia• 2026

Related benchmarks

TaskDatasetResultRank
KnowledgeMMLU-Pro
Score72.5
98
MathematicsAIME 2024
Accuracy77.5
87
MathAIME 2025
Accuracy (%)69.9
26
ScienceGPQA Diamond
Accuracy67
19
Code GenerationCodeForces
ELO1.67e+3
4
Instruction FollowingIFBench
Loose Prompt Success Rate72.2
4
Showing 6 of 6 rows

Other info

Follow for update