Haoran Deng

Papers on Lattice

Total citations

Topics

h-index

Research focus

Robotics & Embodied AI (1)Tool Use & Agents (1)Training Efficiency & Optimization (1)

Frequent co-authors

Xiaoxuan Wang (1)Haixin Wang (1)Ruoyan Li (1)Ruoyan Li (1)

Papers (1)

Feb 25, 2026

Feb 25, 2026·also UCLA

ARLArena: A Unified Framework for Stable Agentic Reinforcement Learning

ARLArena reveals the hidden instability of agentic RL, offering a path to more reliable LLM-based agents via a novel stable policy optimization method (SAMPO).

Xiaoxuan Wang, Haixin Wang, Ruoyan Li +10

Robotics & Embodied AI Tool Use & Agents Training Efficiency & Optimization

Search

Haoran Deng

Research focus

Frequent co-authors

Papers (1)