用大模型模拟反事实场景，解释多智能体行为决策。
Integrating Counterfactual Simulations with Language Models for Explaining Multi-Agent Behaviour
- 让大模型通过'如果'、'移除'等提问，反复试探环境来生成解释。
- 在10个自动驾驶场景中，解释正确率提升至少7.7%，目标预测准确率提高23%。
- 适合关注可解释性与人机信任的自动驾驶研究者。
自主多智能体系统（MAS）可用于自动化复杂任务，但因协调失误或目标错位等问题引发信任担忧。可解释性对用户信任至关重要，但现有方法受限于复杂环境、人为因素及缺乏标准化评估。本文提出基于反事实效应大小模型与大语言模型的智能体解释框架AXIS，通过让大模型以'如果'、'移除'等提示语向环境模拟器发问，多轮观察并整合反事实信息，生成以人为本的行为解释。我们在十种自动驾驶场景下对五种大模型进行评估，采用结合鲁棒性、主观偏好、正确性及目标/动作预测的综合方法，并以外部大模型为评价器。相比基线，AXIS在所有模型上解释正确率提升至少7.7%，四类模型目标预测准确率提升23%，动作预测准确率相当，整体得分最高。代码已开源：https://github.com/gyevnarb/axis。
原文摘要 · Abstract (English)
Autonomous multi-agent systems (MAS) are useful for automating complex tasks but raise trust concerns due to risks such as miscoordination or goal misalignment. Explainability is vital for users' trust calibration, but explainable MAS face challenges due to complex environments, the human factor, and non-standardised evaluation. Leveraging the counterfactual effect size model and LLMs, we propose Agentic eXplanations via Interrogative Simulation (AXIS). AXIS generates human-centred action explanations for multi-agent policies by having an LLM interrogate an environment simulator using prompts like 'whatif' and 'remove' to observe and synthesise counterfactual information over multiple rounds. We evaluate AXIS on autonomous driving across ten scenarios for five LLMs with a comprehensive methodology combining robustness, subjective preference, correctness, and goal/action prediction with an external LLM as evaluator. Compared to baselines, AXIS improves perceived explanation correctness by at least 7.7% across all models and goal prediction accuracy by 23% for four models, with comparable action prediction accuracy, achieving the highest scores overall. Our code is open-sourced at https://github.com/gyevnarb/axis.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。