evoscm / EvoSCM

# EvoSCM: Scientific Belief Revision Through Causal Model Evolution and Experimentation
## 📝 Abstract
Scientific discovery depends on the ability to form hypotheses, test them through experiments, and revise them when evidence disagrees. Existing LLM agents support this process by improving their reasoning or actions, but their scientific beliefs are often scattered across free-form reasoning and difficult to update coherently. This makes it difficult to identify what failed, what should change, and whether revisions remain consistent with prior evidence. We introduce EvoSCM, which represents scientific beliefs as a population of structural causal model (SCM) hypotheses that can be tested and revised across experiments. EvoSCM formulates scientific discovery as a closed loop in which causal hypotheses guide experimentation and experimental outcomes drive causal model evolution. Competing SCM hypotheses make falsifiable predictions and guide discriminative experiments that separate alternative explanations. When observations contradict these predictions, EvoSCM distills discrepancies into correction rules identifying which aspects of the hypotheses fail to explain the evidence. These rules guide revisions to causal dependencies, latent factors, mechanisms, and parameters. Revised hypotheses are validated against accumulated evidence and carried forward to guide subsequent experiments, allowing scientific beliefs to evolve cumulatively. We evaluate EvoSCM across physics, chemistry and materials, and biology. It consistently outperforms baseline agents and existing evolution methods, yielding more accurate explanations and predictions with more effective use of experimental budgets. The evolved SCMs also transfer across base models, suggesting reusable scientific knowledge beyond any single model's reasoning process.
## 🧠 Method Overview
**Figure 2: Overview of EvoSCM.** EvoSCM maintains a population of competing SCM hypotheses and evolves it through a closed discovery loop: (1) *Causal Experiment Design:* the agent abductively interprets accumulated evidence, selects discriminative interventions, and commits to falsifiable predictions tested through experimentation; (2) *SCM Hypothesis Revision:* prediction–observation discrepancies are inductively distilled into correction rules that guide targeted revisions to causal structure and mechanisms, while revised hypotheses are deductively validated against accumulated evidence and structural consistency to form the next-round population.
## ⚙️ Installation and API Configuration
Run the following commands from the repository root:
 [code] 
The simulator can run on CPU; a local GPU is not required for hosted LLM APIs. The smoke test runs all supported simulators without making model requests.
 [code] 
`.env.example` documents the variables; scripts read exported environment variables and do not automatically load `.env`.
Discovery and explanation judging make API requests. Generated scientific Python run