# Do Two AI Scientists Agree?

arXiv: https://arxiv.org/abs/2504.02822 · HTML: https://mlsanity.com/paper/2504.02822/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 48 references that are in the map.
- First posted on arXiv 3 Apr 2025.

## Builds on (most cited first, 20 of 20)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Score-Based Generative Modeling through Stochastic Differential Equations](https://mlsanity.com/paper/2011.13456/) (2020)
- [Deep Unsupervised Learning using Nonequilibrium Thermodynamics](https://mlsanity.com/paper/1503.03585/) (2015)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Generative Modeling by Estimating Gradients of the Data Distribution](https://mlsanity.com/paper/1907.05600/) (2019)
- [Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification](https://mlsanity.com/paper/1502.01852/) (2015)
- [The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery](https://mlsanity.com/paper/2408.06292/) (2024)
- [Hamiltonian Neural Networks](https://mlsanity.com/paper/1906.01563/) (2019)
- [AI Feynman: a Physics-Inspired Method for Symbolic Regression](https://mlsanity.com/paper/1905.11481/) (2019)
- [Discovering Symbolic Models from Deep Learning with Inductive Biases](https://mlsanity.com/paper/2006.11287/) (2020)
- [Learning and Generalization in Overparameterized Neural Networks, Going Beyond Two Layers](https://mlsanity.com/paper/1811.04918/) (2018)
- [Ab-Initio Solution of the Many-Electron Schr"odinger Equation with Deep Neural Networks](https://mlsanity.com/paper/1909.02487/) (2019)
- [Hamiltonian Generative Networks](https://mlsanity.com/paper/1909.13789/) (2019)
- [Simplifying Hamiltonian and Lagrangian Neural Networks via Explicit Constraints](https://mlsanity.com/paper/2010.13581/) (2020)
- [Machine-learning hidden symmetries](https://mlsanity.com/paper/2109.09721/) (2021)
- [LagNetViP: A Lagrangian Neural Network for Video Prediction](https://mlsanity.com/paper/2010.12932/) (2020)
- [Learning Articulated Rigid Body Dynamics with Lagrangian Graph Neural Network](https://mlsanity.com/paper/2209.11588/) (2022)
- [Learning the Dynamics of Particle-based Systems with Lagrangian Graph Neural Networks](https://mlsanity.com/paper/2209.01476/) (2022)

## Built on by (most cited first, 1 of 1)

- [BACON: A fully explainable AI model with graded logic for decision making problems](https://mlsanity.com/paper/2505.14510/) (2025)

## Abstract (arXiv)

When two AI models are trained on the same scientific task, do they learn the same theory or two different theories? Throughout history of science, we have witnessed the rise and fall of theories driven by experimental validation or falsification: many theories may co-exist when experimental data is lacking, but the space of survived theories become more constrained with more experimental data becoming available. We show the same story is true for AI scientists. With increasingly more systems provided in training data, AI scientists tend to converge in the theories they learned, although sometimes they form distinct groups corresponding to different theories. To mechanistically interpret what theories AI scientists learn and quantify their agreement, we propose MASS, Hamiltonian-Lagrangian neural networks as AI Scientists, trained on standard problems in physics, aggregating training results across many seeds simulating the different configurations of AI scientists. Our findings suggests for AI scientists switch from learning a Hamiltonian theory in simple setups to a Lagrangian formulation when more complex systems are introduced. We also observe strong seed dependence of the training dynamics and final learned weights, controlling the rise and fall of relevant theories. We finally demonstrate that not only can our neural networks aid interpretability, it can also be applied to higher dimensional problems.