sbi

Name: sbi
Availability: InStock
Author: smestern

Estimate biophysical model parameters from observed data using simulation-based inference.

1|1|Updated Feb 10, 2026

Search Tags:#parameter estimation #Bayesian inference #simulation-based inference #biophysical model #neural density estimation

Ready to run

npx skills add https://github.com/smestern/sciagent --skill sbi

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: sbi
Source: https://github.com/smestern/sciagent/tree/main/docs/domains/computational-neuro/skills/sbi
Command: npx skills add https://github.com/smestern/sciagent --skill sbi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires torch, numpy, sbi, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill simplifies the process of estimating biophysical model parameters using simulation-based inference (SBI), making complex inference tasks more accessible.

Core Features & Use Cases

Parameter Estimation: Estimate model parameters from observed data without an explicit likelihood.
Bayesian Inference: Offers Bayesian inference capabilities for parameter fitting.
Neural Density Estimation: Utilizes neural networks for density estimation (SNPE, SNLE, SNRE).
Posterior Analysis: Provides tools for posterior analysis and validation.
Use Case: Ideal for researchers in computational neuroscience who need to fit biophysical models to experimental data.

Quick Start

Use the sbi skill to infer HH conductances from a voltage trace.

Frequently Asked Questions about sbi

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I estimate biophysical model parameters without an explicit likelihood function?▼

You can estimate biophysical model parameters without explicit likelihoods by applying simulation-based inference, which trains neural density estimators on simulated data. The Skill uses SNPE, SNLE, and SNRE algorithms for posterior analysis.

What is simulation-based inference and how does neural density estimation work?▼

Simulation-based inference uses neural density estimation to approximate posterior distributions without explicit likelihoods. Neural networks learn the relationship between simulated data and parameters, enabling Bayesian inference for complex biophysical models.

Can I use this for fitting computational neuroscience models to experimental voltage traces?▼

Yes, simulation-based inference is ideal for computational neuroscience research fitting biophysical models to experimental data. You can infer parameters such as HH conductances directly from observed voltage traces using neural density estimation techniques.

Do I need PyTorch to run simulation-based inference for parameter estimation?▼

Yes, PyTorch is required because neural density estimation relies on its deep learning framework. You need PyTorch, NumPy, and the sbi library installed to perform simulation-based inference and posterior analysis.

What is the best way to validate posterior distributions from neural density estimation?▼

The best way to validate posterior distributions is by using the built-in posterior analysis tools provided by the simulation-based inference framework. These tools evaluate the accuracy and reliability of neural density estimation results.

When should I choose SNPE over SNLE or SNRE for Bayesian inference?▼

SNPE, SNLE, and SNRE offer distinct neural density estimation approaches for simulation-based inference. SNPE directly approximates the posterior, while SNLE and SNRE estimate likelihoods or ratios, providing flexibility for different Bayesian inference scenarios.