cs.LGNov 9, 2026

Evidence or Marginal? What Moment-Vector Propagation Computes in Polynomial Density Neurons

Authors: Przemysław Spyra

Organizations: AGH University of Krakow

Abstract

When a neuron's input is lost, or known through a belief qq, what should it output? For a neuron storing a joint polynomial density NN with input density DD, two formulas compete. pCD=Eq[N]/Eq[D]p_\mathrm{CD}=E_q[N]/E_q[D] costs one matrix-vector product; Hierarchical Correlation Reconstruction introduces it as a constant-denominator approximation to pEU=Eq[N/D]p_\mathrm{EU}=E_q[N/D]. For a density model both are exact answers to different questions: pCDp_\mathrm{CD} is Bayesian updating with the belief as virtual evidence; pEUp_\mathrm{EU} installs the belief as a new marginal (Jeffrey's rule). With a missing input pCDp_\mathrm{CD} returns the model's conditional; pEUp_\mathrm{EU} generally does not. This post hoc reading is proved, and illustrated outside the pair family up to k=8k=8 inputs; it has no pre-registered test, only one exploratory real-data pilot. Post hoc, closed forms fall within a factor 2 of all 112 measured gaps, and negative density begins near input correlation 1/k1/k. In a pre-registered synthetic study (up to 32 inputs), 5 of 15 hypotheses are supported, 2 inconclusive and 8 not supported, including all three on transfer beyond equicorrelated Gaussians. Post hoc, where pEUp_\mathrm{EU} is defined, truncation usually outweighs the gap for Beta beliefs in pair neurons; triple neurons under flat and data beliefs are exceptions.

Figures & tables

Appendix figures & tables38 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Same Answer, Different Future: Dissociating Belief Content and Evidence Weight in Language Models

    Nov 5, 2026Weihan Li, Junhao Wu, Xinlei Chen +1Belief Updating in LLMsProbability Elicitation from Language Models