PPL
Sung Ju Hwang
Papers (6)
Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use AgentsAgent Explorative Policy Optimization for Multimodal Agentic ReasoningHINT-SD: Targeted Hindsight Self-Distillation for Long-Horizon AgentsNudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVRSAGE: Shaping Anchors for Guided Exploration in RLVR of LLMsPREPING: Building Agent Memory without Tasks