Scenario 2: Data Analysis and Neural Decoding

A comprehensive tutorial on experimental neural data analysis, topological decoding, and RNN dynamics.

Tutorial List

Experimental Data Analysis

RNN Dynamics Analysis

Tutorial Overview

ASA Pipeline Tutorial

Covers the complete pipeline from spike/x/y/t input to TDA, decoding, CohoMap/CohoSpace/PathCompare, and FR/FRM, with corresponding example scripts in the repository.

1D CANN ROI Bump Fitting Tutorial

Demonstrates how to use roi_bump_fits to extract bump parameters and generate animations for analyzing ring attractor dynamics.

Cell Classification Tutorial

Illustrates a cell classification workflow based on GridScore and autocorrelation features, including examples of single-cell scoring and grid module segmentation.

RNN Fixed-Point Analysis Tutorial (FlipFlop Task)

This tutorial provides a detailed guide on using the FixedPointFinder tool to analyze the dynamical properties of recurrent neural networks (RNNs):

  • Theoretical Foundation: Understanding the concept of fixed points in dynamical systems

  • FlipFlop Task: Training an RNN to perform a multi-channel memory task

  • Fixed-Point Identification: Using optimization methods to locate stable and unstable fixed points

  • Visualization Analysis: Displaying fixed-point distributions in state space via PCA dimensionality reduction

  • Multi-Configuration Comparison: Comparing fixed-point structures across 2-bit, 3-bit, and 4-bit tasks

Key Finding: For an N-bit FlipFlop task, a successfully trained RNN learns to create 2^N stable fixed points—each corresponding to a unique combination of memory states.

Example Code

  • examples/experimental_data_analysis: Scripts related to the ASA pipeline (TDA/decoding/CohoMap/CohoSpace/PathCompare/FR, etc.)

  • examples/cell_classification: Examples for cell classification and related analyses