ALFWorld

ALFWorld is a dataset/benchmark that evaluates the performance of large language models (LLMs) as autonomous agents in multi-turn tasks by providing a structured environment for reinforcement learning.

Papers using ALFWorld (55)

RLVMR: Reinforcement Learning with Verifiable Meta-Reasoning Rewards for Robust Long-Horizon Agents2025 · 35 citesGTR: Guided Thought Reinforcement Prevents Thought Collapse in RL-based VLM Agent Training2025 · 11 citesSEEA-R1: Tree-Structured Reinforcement Fine-Tuning for Self-Evolving Embodied Agents2025 · 10 citesRetrospex: Language Agent Meets Offline Reinforcement Learning Critic2025 · 7 citesBeyond Policy Optimization: A Data Curation Flywheel for Sparse-Reward Long-Horizon Planning2025 · 3 citesWhen Does Muon Help Agentic Reinforcement Learning?20263SPO: State-Score-Supervised Policy Optimization for LLM Agents2026Semantic Consistency Policy Optimization for Reinforcement Learning of LLM Agents2026Hera: Learning Long-Horizon Coordination for Device-Cloud Collaborative LLM Agents2026StepOPSD: Step-Aware Online Preference Distillation for Agent Reinforcement Learning2026Graph-Enhanced Policy Optimization in LLM Agent Training2025 · 2 citesLearn the Ropes, Then Trust the Wins: Self-imitation with Progressive Exploration for Agentic Reinforcement Learning2025 · 1 citesGroup-in-Group Policy Optimization for LLM Agent Training2025 · 1 citesProcess Reward Models for LLM Agents: Practical Framework and Directions2025 · 1 citesSIRI: Self-Internalizing Reinforcement Learning with Intrinsic Skills for LLM Agent Training2026Policy-Conditioned Counterfactual Credit for Verifiable Reinforcement Learning of Long-Horizon Language Agents2026When Denser Credit Is Not Enough: Evidence-Calibrated Policy Optimization for Long-Horizon LLM Agent Training2026T$^2$PO: Uncertainty-Guided Exploration Control for Stable Multi-Turn Agentic Reinforcement Learning2026Selective Rollout: Mid-Trajectory Termination for Multi-Sample Agent RL2026Milestone-Guided Policy Learning for Long-Horizon Language Agents2026Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning2026StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction2026Evolving-RL: End-to-End Optimization of Experience-Driven Self-Evolving Capability within Agents2026Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning2026GAGPO: Generalized Advantage Grouped Policy Optimization2026Self-Distilled Agentic Reinforcement Learning2026SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization2026Hierarchical Reinforcement Learning with Augmented Step-Level Transitions for LLM Agents2026Agent^2 RL-Bench: Can LLM Agents Engineer Agentic RL Post-Training?2026HiMAC: Hierarchical Macro-Micro Learning for Long-Horizon LLM Agents2026Hindsight Credit Assignment for Long-Horizon LLM Agents2026Dynamic Dual-Granularity Skill Bank for Agentic RL2026RLAnything: Forge Environment, Policy, and Reward Model in Completely Dynamic RL System2026SkillRL: Evolving Agents via Recursive Skill-Augmented Reinforcement Learning2026Think Fast and Slow: Step-Level Cognitive Depth Adaptation for LLM Agents2026HiPER: Hierarchical Reinforcement Learning with Explicit Credit Assignment for Large Language Model Agents2026Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Tasks2026Paying Less Generalization Tax: A Cross-Domain Generalization Study of RL Training for LLM Agents2026Differentiable Evolutionary Reinforcement Learning2025A Practitioner's Guide to Multi-turn Agentic Reinforcement Learning2025Dyna-Mind: Learning to Simulate from Experience for Better AI Agents2025SALT: Step-level Advantage Assignment for Long-horizon Agents via Trajectory Graph2025DeepAgent: A General Reasoning Agent with Scalable Toolsets2025Code Driven Planning with Domain-Adaptive Critic2025Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM Agents2025EPO: Entropy-regularized Policy Optimization for LLM Agents Reinforcement Learning2025Enhancing Vision-Language Model Training with Reinforcement Learning in Synthetic Worlds for Real-World Success2025Enhancing Decision-Making of Large Language Models via Actor-Critic2025Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning2025Cache-Efficient Posterior Sampling for Reinforcement Learning with LLM-Derived Priors Across Discrete and Continuous Domains2025Divide and Conquer: Grounding LLMs as Efficient Decision-Making Agents via Offline Hierarchical Reinforcement Learning2025SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution2025GFlowVLM: Enhancing Multi-step Reasoning in Vision-Language Models with Generative Flow Networks2025ReAct Meets ActRe: When Language Agents Enjoy Training Data Autonomy2024 · 1 citesOffline Reinforcement Learning for LLM Multi-Step Reasoning2024 · 1 cites