CosmoSlider: An educational tool for cosmology

TL;DR

CosmoSlider tackles the challenge of interactively teaching how cosmological parameters shape the CMB power spectra by introducing a lightweight neural-network emulator built with TensorFlow Lite, enabling real-time, multi-parameter exploration within the $\Lambda$CDM framework. Available as both an iOS app and a web tool, it renders $C_\ell$ spectra (TT, TE, EE, $\phi\phi$) in real time, using the first $100$ multipoles and cubic interpolation for the rest to maintain responsiveness, and supports overlaying Planck data to guide parameter selection. The tool aims to enhance education by providing immediate visual feedback and a hands-on way to build intuition about CMB physics, while clearly complementing traditional EB solver-based instruction. Planned updates promise broader observables and datasets, stronger customization, and expanded data integrations, reinforcing CosmoSlider as a scalable, cross-platform educational resource for teaching cosmology.

Abstract

Understanding how cosmological parameters influence the cosmic microwave background (CMB) power spectra is a central component of modern cosmology education, but interactive exploration is often limited by computational cost or technical complexity. We present CosmoSlider, a lightweight visualization tool that enables real-time exploration of CMB power spectra as multiple cosmological parameters are varied simultaneously. The tool employs a neural-network emulator implemented using TensorFlow Lite, allowing rapid evaluation of spectra without relying on large grids of precomputed models or on-demand execution of Einstein--Boltzmann solvers. CosmoSlider is available both as an iOS application and as a web-based tool, making it accessible across platforms and suitable for use in classrooms, lectures, and self-guided study. By providing immediate visual feedback, CosmoSlider supports the development of intuition for the physical processes underlying CMB anisotropies and serves as a complementary resource to traditional theoretical instruction.

Figures (4)

• Figure 1: The CMB power spectra computed by class using the Planck best-fit parameters Planck:2018vyg along with the observed data points from Planck.
• Figure 2: Amount of RAM needed to store $M^N$ precomputed spectra of 400 bytes as a function of the resolution, $M$, along each of $N=6$ dimensions. The dashed lines indicate the RAM needed specifically for resolutions of 10 and 100.
• Figure 3: Visual representation of the two versions of CosmoSlider. On the left, the iOS version is shown on phone screens, and on the right, the web version is shown on a computer screen. Both versions have dark and light appearance modes.
• Figure 4: Flowchart of importing a new neural network model into the iOS version of CosmoSlider. The structure of the models read by the application is also depicted.