# Learning to Outgrow a Theory: Experimental Discovery Beyond the Initial Hypothesis Space

arXiv: https://arxiv.org/abs/2610.07627 · HTML: https://mlsanity.com/paper/2610.07627/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 32 references that are in the map.
- First posted on arXiv 6 Oct 2026.

## Builds on (most cited first, 15 of 15)

- [The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search](https://mlsanity.com/paper/2504.08066/) (2025)
- [LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models](https://mlsanity.com/paper/2504.10415/) (2025)
- [Interpretable Machine Learning for Science with PySR and SymbolicRegression.jl](https://mlsanity.com/paper/2305.01582/) (2023)
- [Automated Hypothesis Validation with Agentic Sequential Falsifications](https://mlsanity.com/paper/2502.09858/) (2025)
- [A Probabilistic Framework for LLM-Based Model Discovery](https://mlsanity.com/paper/2602.18266/) (2026)
- [LLM-AutoSciLab: Closed-Loop Scientific Discovery via Active Experimentation with LLMs](https://mlsanity.com/paper/2605.24043/) (2026)
- [DiscoverPhysics: Benchmarking LLMs for Out-of-the-Box Scientific Thinking](https://mlsanity.com/paper/2605.26087/) (2026)
- [LLM-ACES: Closed-Loop Discovery of Dynamical Systems with LLM-Guided Adaptive Search](https://mlsanity.com/paper/2606.25039/) (2026)
- [ATLAS: Active Theory Learning for Automated Science](https://mlsanity.com/paper/2606.12386/) (2026)
- [Are all models wrong? Fundamental limits in distribution-free empirical model falsification](https://mlsanity.com/paper/2502.06765/) (2025)
- [Representative, Informative, and De-Amplifying: Requirements for Robust Bayesian Active Learning under Model Misspecification](https://mlsanity.com/paper/2506.07805/) (2025)
- [Robust Simulation Based Inference](https://mlsanity.com/paper/2508.02404/) (2025)
- [EvoSCM: Scientific Belief Revision Through Causal Model Evolution and Experimentation](https://mlsanity.com/paper/2609.01526/) (2026)
- [Evidence-Informed LLM Beliefs for Continual Scientific Discovery](https://mlsanity.com/paper/2606.29182/) (2026)
- [Adversarial Causal Intervention Falsification](https://mlsanity.com/paper/2608.06427/) (2026)

## Abstract (arXiv)

Scientific discovery systems typically optimize experiments within a fixed hypothesis space. This creates a failure mode when all available candidates omit the same missing mechanism: candidate disagreement can collapse even while the model class is systematically wrong. We formulate experimental model-class revision, in which a discovery policy jointly proposes a structural edit and a diagnostic experiment that tests whether that edit is necessary. The method couples a class-level distinguishability objective, in which one shared parameterization must explain all selected experiments, with anytime-valid sequential evidence that triggers structural revision only after the current class is rejected. On 400 held-out controlled dynamical environments, the joint policy reaches 89.5% exact recovery with a budget of 32 real experiments, improving the strongest matched baseline by 10.0 percentage points while requiring fewer executed experiments and candidate fits. The learned revision-experiment pairing transfers across unseen mechanism combinations, held-out but expressible primitives, parameter extrapolation, and shifted experiment costs; when the true mechanism is outside the edit grammar, it detects library insufficiency in 88% of cases with a 5.5% false-support rate. Revision gains also transfer to ODEBench and ODEBase model-library tasks, as well as DiscoverPhysics worlds. These results support a view of scientific discovery in which deciding what mechanisms a theory should make expressible and where to collect evidence are treated as a single sequential decision problem.