Discovering Mechanistic Models of Neural Activity: System Identification in an in Silico Zebrafish

2026-02-04

AI Summary

This study addresses the lack of verifiable ground truth in neuroscience for evaluating mechanistic models by introducing a whole-brain neuromechanical simulation platform based on larval zebrafish, which provides a transparent and controllable benchmark environment. The work innovatively integrates large language model (LLM)-guided symbolic regression with neural architectural priors to automatically discover predictive and interpretable models of neural mechanisms through tree search. The resulting models significantly outperform conventional baselines, demonstrate strong out-of-distribution generalization, and accurately recapitulate key functional architectures observed in real neural circuits.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Constructing mechanistic models of neural circuits is a fundamental goal of neuroscience, yet verifying such models is limited by the lack of ground truth. To rigorously test model discovery, we establish an in silico testbed using neuromechanical simulations of a larval zebrafish as a transparent ground truth. We find that LLM-based tree search autonomously discovers predictive models that significantly outperform established forecasting baselines. Conditioning on sensory drive is necessary but not sufficient for faithful system identification, as models exploit statistical shortcuts. Structural priors prove essential for enabling robust out-of-distribution generalization and recovery of interpretable mechanistic models. Our insights provide guidance for modeling real-world neural recordings and offer a broader template for AI-driven scientific discovery.

Problem

• mechanistic models
• system identification
• neural circuits
• ground truth
• model discovery

Innovation

• mechanistic modeling
• system identification
• LLM-based tree search
• structural priors
• in silico neuroscience