Universal Beta Incidence Angles: Cauchy Rigidity and Infinite Arrangements
arXiv:2609.00603
2026
Architecture
2 ideas extracted · analyzed Sep 2, 2026
What the math gives to ML
The paper gives an unusually strong distributional invariance for a deterministic reciprocal-projection construction: any simplex-weighted collection of nonzero directions produces an output whose squared cosine with the input is exactly Beta(1/2,(p-1)/2), regardless of the directions, rank, or overcompleteness. This can be transferred into neural networks as a geometry-aware angular mixing layer or as a calibration target for representation maps, providing controlled angular diversity without injecting random output directions. The main engineering risk is the pole at a_j^T U=0, so practical implementations must use clipping or a smooth reciprocal and explicitly measure how much the exact beta law degrades.
Ideas from this paper
Unverified
2026
Insert a projection-space mixer that combines several fixed or learned directions using reciprocal correlations with the current feature, then normalize the result. The exact construction has a universal beta law for its squared input-output cosine, so it can create controlled angular diversity while remaining deterministic and independent of the chosen direction dictionary.
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Use the reciprocal arrangement as a probe of whether a learned representation has the intended angular response, and penalize deviations from the paper's universal beta distribution. This converts the theorem into a distribution-level regularizer rather than assuming that the reciprocal layer itself improves task loss.
Useful5/10
Difficulty4/10
Novelty7/10