AppWorld

The 'AppWorld' dataset/benchmark contains a collection of applications and their associated contexts, used to evaluate the ability of language agents to ground user instructions in the relevant environmental information.

Papers using AppWorld (15)

From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flaws2026ACCORD: Action-Conditioned Contextual Grounding for Language Agents2026Hera: Learning Long-Horizon Coordination for Device-Cloud Collaborative LLM Agents2026ExpGraph: Model-Agnostic Experience Learning with Graph-Structured Memory for LLM Agents2026Pushing the Limits of LLM Tool Calling via Experiential Knowledge Integration and Activation2026Metis: Bridging Text and Code Memory for Self-Evolving Agents2026Keep Policy Gradient in Charge: Sibling-Guided Credit Distillation for Long-Horizon Tool-Use Agents2026Learning and Reusing Policy Decompositions for Hierarchical Generalized Planning with LLM Agents2026HINT-SD: Targeted Hindsight Self-Distillation for Long-Horizon Agents2026CoEvolve: Training LLM Agents via Agent-Data Mutual Evolution2026Three Roles, One Model: Role Orchestration At Inference Time To Close The Performance Gap Between Small And Large Agents2026Toward Scalable Verifiable Reward: Proxy State-Based Evaluation for Multi-turn Tool-Calling LLM Agents2026Reinforcement Learning for Self-Improving Agent with Skill Library2025ACON: Optimizing Context Compression for Long-horizon LLM Agents2025Prost: Progressive Sub-task Training For Pareto-optimal Multi-agent Systems Using Small Language Models2025