ScienceWorld

ScienceWorld is a benchmark dataset used to evaluate the performance and generalization of large language model agents in complex interactive decision-making tasks.

Papers using ScienceWorld (7)

Exploratory Memory-Augmented LLM Agent via Hybrid On- and Off-Policy Optimization2026Blueprint First, Model Second: A Framework for Deterministic LLM Workflow2025Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for LLM Agents2026Think Fast and Slow: Step-Level Cognitive Depth Adaptation for LLM Agents2026SkillGen: Learning Domain Skills for In-Context Sequential Decision Making2025KnowMap: Efficient Knowledge-Driven Task Adaptation for LLMs2025Retrospex: Language Agent Meets Offline Reinforcement Learning Critic2025