Generating Causal Explanations of Vehicular Agent Behavioural Interactions with Learnt Reward Profiles

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Howard, Rhys, Hawes, Nick, Kunze, Lars
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918202647248896
author Howard, Rhys
Hawes, Nick
Kunze, Lars
author_facet Howard, Rhys
Hawes, Nick
Kunze, Lars
contents Transparency and explainability are important features that responsible autonomous vehicles should possess, particularly when interacting with humans, and causal reasoning offers a strong basis to provide these qualities. However, even if one assumes agents act to maximise some concept of reward, it is difficult to make accurate causal inferences of agent planning without capturing what is of importance to the agent. Thus our work aims to learn a weighting of reward metrics for agents such that explanations for agent interactions can be causally inferred. We validate our approach quantitatively and qualitatively across three real-world driving datasets, demonstrating a functional improvement over previous methods and competitive performance across evaluation metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2503_14557
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Generating Causal Explanations of Vehicular Agent Behavioural Interactions with Learnt Reward Profiles
Howard, Rhys
Hawes, Nick
Kunze, Lars
Artificial Intelligence
Multiagent Systems
Robotics
I.2.0; I.2.6; I.2.9; I.2.11; I.6.0
Transparency and explainability are important features that responsible autonomous vehicles should possess, particularly when interacting with humans, and causal reasoning offers a strong basis to provide these qualities. However, even if one assumes agents act to maximise some concept of reward, it is difficult to make accurate causal inferences of agent planning without capturing what is of importance to the agent. Thus our work aims to learn a weighting of reward metrics for agents such that explanations for agent interactions can be causally inferred. We validate our approach quantitatively and qualitatively across three real-world driving datasets, demonstrating a functional improvement over previous methods and competitive performance across evaluation metrics.
title Generating Causal Explanations of Vehicular Agent Behavioural Interactions with Learnt Reward Profiles
topic Artificial Intelligence
Multiagent Systems
Robotics
I.2.0; I.2.6; I.2.9; I.2.11; I.6.0
url https://arxiv.org/abs/2503.14557