Haoyu Wang

UltraHorizon: Benchmarking LLM-Agent Capabilities in Ultra Long-Horizon Scenarios

A Simple "Motivation" Can Enhance Reinforcement Finetuning of Large Reasoning Models

Rethinking the Role of Dynamic Sparse Training for Scalable Deep Reinforcement Learning

Safety Reasoning with Guidelines

Lifelong Safety Alignment for Language Models

Mastering Massive Multi-Task Reinforcement Learning via Mixture-of-Expert Decision Transformer