AI RESEARCH
Super Agents and Confounders: Influence of surrounding agents on vehicle trajectory prediction
arXiv CS.LG
•
ArXi:2604.03463v1 Announce Type: new In highly interactive driving scenes, trajectory prediction is conditioned on information from surrounding traffic participants such as cars and pedestrians. Our main contribution is a comprehensive analysis of state-of-the-art trajectory predictors, which reveals a surprising and critical flaw: many surrounding agents degrade prediction accuracy rather than improve it. Using Shapley-based attribution, we rigorously nstrate that models runs.