Salvato in:
Dettagli Bibliografici
Autori principali: Koblischke, Nolan, Jang, Hyunseok, Menou, Kristen, Ali-Dib, Mohamad
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2501.18411
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912400342515712
author Koblischke, Nolan
Jang, Hyunseok
Menou, Kristen
Ali-Dib, Mohamad
author_facet Koblischke, Nolan
Jang, Hyunseok
Menou, Kristen
Ali-Dib, Mohamad
contents Modern science emerged from reasoning over repeatedly-observed planetary motions. We present Gravity-Bench-v1, an environment-based benchmark that challenges AI agents on tasks that parallel this historical development. Gravity-Bench-v1 evaluates agents on the discovery of physics concealed within a dynamic environment, using rigorous gravitational dynamics simulations. Gravity-Bench includes out-of-distribution cases, i.e. with physics that deviates from the real world, to evaluate true scientific generalization capabilities. Agents must plan to collect data within an experimental budget and must perform a dynamic form of data analysis and reasoning to solve tasks efficiently. Our benchmark admits an open-ended space of solutions. Reference solutions for each task are provided to calibrate AI performance against human expertise. Technically at an upper-undergraduate level, our benchmark proves challenging to baseline AI agents. Gravity-Bench-v1 and planned extensions should help map out AI progress towards scientific discovery capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18411
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Gravity-Bench-v1: A Benchmark on Gravitational Physics Discovery for Agents
Koblischke, Nolan
Jang, Hyunseok
Menou, Kristen
Ali-Dib, Mohamad
Artificial Intelligence
Instrumentation and Methods for Astrophysics
Computational Physics
Modern science emerged from reasoning over repeatedly-observed planetary motions. We present Gravity-Bench-v1, an environment-based benchmark that challenges AI agents on tasks that parallel this historical development. Gravity-Bench-v1 evaluates agents on the discovery of physics concealed within a dynamic environment, using rigorous gravitational dynamics simulations. Gravity-Bench includes out-of-distribution cases, i.e. with physics that deviates from the real world, to evaluate true scientific generalization capabilities. Agents must plan to collect data within an experimental budget and must perform a dynamic form of data analysis and reasoning to solve tasks efficiently. Our benchmark admits an open-ended space of solutions. Reference solutions for each task are provided to calibrate AI performance against human expertise. Technically at an upper-undergraduate level, our benchmark proves challenging to baseline AI agents. Gravity-Bench-v1 and planned extensions should help map out AI progress towards scientific discovery capabilities.
title Gravity-Bench-v1: A Benchmark on Gravitational Physics Discovery for Agents
topic Artificial Intelligence
Instrumentation and Methods for Astrophysics
Computational Physics
url https://arxiv.org/abs/2501.18411