Sampling-based Causal Inference in Cue Combination and its Neural Implementation

TL;DR

This paper introduces a hierarchical importance sampling algorithm for causal inference in cue combination and demonstrates its neural implementation, which is more realistic and generalizable than previous methods.

Contribution

It proposes a novel importance sampling-based inference algorithm and a neural circuit model that can be applied to various causal inference problems.

Findings

Algorithm converges to accurate values with increasing samples

Neural circuit implementation is simple and generalizable

Applicable to multi-stimuli cause inference and same-different judgment

Abstract

Causal inference in cue combination is to decide whether the cues have a single cause or multiple causes. Although the Bayesian causal inference model explains the problem of causal inference in cue combination successfully, how causal inference in cue combination could be implemented by neural circuits, is unclear. The existing method based on calculating log posterior ratio with variable elimination has the problem of being unrealistic and task-specific. In this paper, we take advantages of the special structure of the Bayesian causal inference model and propose a hierarchical inference algorithm based on importance sampling. A simple neural circuit is designed to implement the proposed inference algorithm. Theoretical analyses and experimental results demonstrate that our algorithm converges to the accurate value as the sample size goes to infinite. Moreover, the neural circuit we design can be easily generalized to implement inference for other problems, such as the multi-stimuli cause inference and the same-different judgment.