AI-Newton: A Concept-Driven Physical Law Discovery System without Prior Physical Knowledge
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914194666815488 |
|---|---|
| author | Fang, You-Le Jian, Dong-Shan Li, Xiang Ma, Yan-Qing |
| author_facet | Fang, You-Le Jian, Dong-Shan Li, Xiang Ma, Yan-Qing |
| contents | While current AI-driven methods excel at deriving empirical models from individual experiments, a significant challenge remains in uncovering the common fundamental physics that underlie these models -- a task at which human physicists are adept. To bridge this gap, we introduce AI-Newton, a novel framework for concept-driven scientific discovery. Our system autonomously derives general physical laws directly from raw, multi-experiment data, operating without supervision or prior physical knowledge. Its core innovations are twofold: (1) proposing interpretable physical concepts to construct laws, and (2) progressively generalizing these laws to broader domains. Applied to a large, noisy dataset of mechanics experiments, AI-Newton successfully rediscovers foundational and universal laws, such as Newton's second law, the conservation of energy, and the universal gravitation. This work represents a significant advance toward autonomous, human-like scientific discovery. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_01538 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | AI-Newton: A Concept-Driven Physical Law Discovery System without Prior Physical Knowledge Fang, You-Le Jian, Dong-Shan Li, Xiang Ma, Yan-Qing Artificial Intelligence Machine Learning Symbolic Computation High Energy Physics - Phenomenology Classical Physics While current AI-driven methods excel at deriving empirical models from individual experiments, a significant challenge remains in uncovering the common fundamental physics that underlie these models -- a task at which human physicists are adept. To bridge this gap, we introduce AI-Newton, a novel framework for concept-driven scientific discovery. Our system autonomously derives general physical laws directly from raw, multi-experiment data, operating without supervision or prior physical knowledge. Its core innovations are twofold: (1) proposing interpretable physical concepts to construct laws, and (2) progressively generalizing these laws to broader domains. Applied to a large, noisy dataset of mechanics experiments, AI-Newton successfully rediscovers foundational and universal laws, such as Newton's second law, the conservation of energy, and the universal gravitation. This work represents a significant advance toward autonomous, human-like scientific discovery. |
| title | AI-Newton: A Concept-Driven Physical Law Discovery System without Prior Physical Knowledge |
| topic | Artificial Intelligence Machine Learning Symbolic Computation High Energy Physics - Phenomenology Classical Physics |
| url | https://arxiv.org/abs/2504.01538 |