Open-Universe Assistance Games
This addresses the challenge for AI agents to interpret evolving human preferences in open-ended environments, representing an incremental advance in goal inference techniques.
The paper tackles the problem of embodied AI agents needing to infer diverse human goals not predefined, by introducing Open-Universe Assistance Games and the GOOD method, which extracts natural language goals from dialogue and outperforms baselines in text-based domains like grocery shopping and simulated robotics.
Embodied AI agents must infer and act in an interpretable way on diverse human goals and preferences that are not predefined. To formalize this setting, we introduce Open-Universe Assistance Games (OU-AGs), a framework where the agent must reason over an unbounded and evolving space of possible goals. In this context, we introduce GOOD (GOals from Open-ended Dialogue), a data-efficient, online method that extracts goals in the form of natural language during an interaction with a human, and infers a distribution over natural language goals. GOOD prompts an LLM to simulate users with different complex intents, using its responses to perform probabilistic inference over candidate goals. This approach enables rich goal representations and uncertainty estimation without requiring large offline datasets. We evaluate GOOD in a text-based grocery shopping domain and in a text-operated simulated household robotics environment (AI2Thor), using synthetic user profiles. Our method outperforms a baseline without explicit goal tracking, as confirmed by both LLM-based and human evaluations.