Documentdetail
ID kaart
oai:arXiv.org:2410.09829
Onderwerp
Computer Science - Computation and... Computer Science - Information Ret... Computer Science - RoboticsAuteur
Rubavicius, Rimvydas Miceli-Barone, Antonio Valerio Lascarides, Alex Ramamoorthy, SubramanianCategorie
Computer Science
Jaar
2024
vermelding datum
16-10-2024
Trefwoorden
usingBeschrijving
Cyber-physical systems like autonomous vehicles are tested in simulation before deployment, using domain-specific programs for scenario specification.
To aid the testing of autonomous vehicles in simulation, we design a natural language interface, using an instruction-following large language model, to assist a non-coding domain expert in synthesising the desired scenarios and vehicle behaviours.
We show that using it to convert utterances to the symbolic program is feasible, despite the very small training dataset.
Human experiments show that dialogue is critical to successful simulation generation, leading to a 4.5 times higher success rate than a generation without engaging in extended conversation.
;Comment: 6 pages, 6 figures, 2 tables
Rubavicius, Rimvydas,Miceli-Barone, Antonio Valerio,Lascarides, Alex,Ramamoorthy, Subramanian, 2024, Generating Driving Simulations via Conversation
Understanding Likelihood Over-optimisation in Direct Alignment Algorithms
algorithms alignment completion human optimisation direct likelihood performance
Neuroscience Engagement and Outreach on Neurodevelopmental Disorders: A Report from India
report current engagement neuroscience public project
