astro-ph.GAMar 31, 2026

Schrödinger's Seed: Purr-fect Initialization for an Impurr-fect Universe

Authors: Mi chen, Renhao Ye

Organizations: Kapteyn Astronomical Institute, University of Groningen, Postbus 800, 9700 AV Groningen, The Netherlands · School of Astronomy and Space Science, University of Chinese Academy of Sciences, 1 East Yanqi Lake Rd., Beijing 100049, P.R. China · Shanghai Astronomical Observatory, Chinese Academy of Sciences, 80 Nandan Rd., Shanghai 200030, China

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Context. Random seed selection in deep learning is often arbitrary -- conventionally fixed to values such as 42, a number with no known feline endorsement. Aims. We propose that cats, as liminal beings with a historically ambiguous relationship to quantum mechanics, are better suited to this task than random integers. Methods. We construct a cat-driven seed generator inspired by the first Friedmann equation, and test it by mapping 21 domestic cats' physical properties -- mass, coat pattern, eye colour, and name entropy -- via a Monte ``Catlo'' sampling procedure. Results. Cat-driven seeds achieve a mean accuracy of 92.58%, outperforming the baseline seed of 42 by ∼\sim2.5%. Cats from astrophysicist households perform marginally better, suggesting cosmic insight may be contagious. Conclusions. The Universe responds better to cats than to arbitrary integers. Whether cats are aware of this remains unknown.

Figures & tables

Figure 1: Cat parameter catalogue. Each card displays (from top to bottom): a portrait photograph; the cat’s name (starred entries indicate cats owned by current or former astrophysicists); sex (M/F), date of birth, and body mass; coat pattern code and eye colour; and a colour-coded accuracy bar. The bar colour indicates performance relative to the baseline seed 42 (accuracy = 0.904): darker forest green denotes higher accuracy, while grey-blue denotes lower accuracy. The seed values are generated using the cat-driven Random Seed Generator with Monte "Catlo", based on each cat’s intrinsic parameters. A control entry with the fixed seed 42 is listed separately at the bottom as a baseline reference.
More in this topic

Neural Network Initialization

Neural network initialization studies how starting weights shape signal propagation, optimization, generalization, and reproducibility.

See all 51 papers

Explore similar work

CardsList

  1. Generating from Discrete Distributions Using Diffusions: Insights from Random Constraint Satisfaction Problems

    Mar 21, 2026Alankrita Bhatt, Mukur Gupta, Germain Kolossov +1Diffusion SamplingDiffusion Models

    Granica Computing Inc. – granica.ai · Stanford University

  2. The FID Lottery: Quantifying Hidden Randomness in Generative-Model Evaluation

    Jun 18, 2026Nicolas Dufour, Alexei A. Efros, Patrick PérezML ReproducibilityImage Generation Evaluation

    Kyutai · UC Berkeley

  3. Schrödinger's Cat: Probabilistic Representation and Prediction of Potential Scene Kinematics

    Jul 28, 2026Timy Phan, Jannik Wiese, Björn OmmerPredictive Representation LearningMotion Planning