BabyAI

'BabyAI' is a dataset/benchmark used to evaluate the reasoning and decision-making abilities of agents, particularly in the context of task completion efficiency.

Papers using BabyAI (4)

Environment-Grounded Automated Prompt Optimization for LLM Game Agents2026SkillGen: Learning Domain Skills for In-Context Sequential Decision Making2025DEPO: Dual-Efficiency Preference Optimization for LLM Agents2025RAG-Modulo: Solving Sequential Tasks using Experience, Critics, and Language Models2024