Working on path finding for directional acyclic graphs. Need some help with list comprehension.

path_finder.py

import networkx as nx

G = nx.DiGraph()
G.add_nodes_from([ i for i in range(0, 5) ])
G.add_edges_from([(0,1), (0,2), (1, 3), (1, 4)])

class Finder:
    def __init__(self, graph):
        self.graph = graph
    
    def find_paths_from(self, start_point):
        neighbors = self.graph.neighbors(start_point) # Find all other nodes start_point can connect to
        if neighbors == []: 
            return [start_point]
        else:
            return [ [start_point] + self.find_paths_from(n) for n in neighbors ]

F = Finder(G)
F.find_paths_from(0) #=> [[0, [1, 3], [1, 4]], [0,2]]
# I want this to return [[0, 1, 3], [0, 1, 4], [0, 2]], How can I do that?

This is from my first paid project. Made $200 and realized I could make money doing this, and now it's 8 years later.