Publications

This page lists papers and technical reports associated with RLinf. Detailed publication pages (with results and quickstart links) are linked below.

Detailed publication pages

• RLinf-USER — Unified system for real-world online policy learning arXiv:2602.07837

• RLinf-VLA — Unified framework for VLA+RL training arXiv:2510.06710

• RLinf-Co — Reinforcement learning-based sim-real co-training for VLA models arXiv:2602.12628

• RLinf — Flexible and efficient RL system arXiv:2509.15965

• πRL — Online RL fine-tuning for flow-based VLA models arXiv:2510.25889

• WoVR — World model-based RL fine-tuning for VLA policies arXiv:2602.13977

• WideSeek-R1 — Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning arXiv:2602.04634