SPICENet Visualization vs Learning Transformer

SpiceNet is an educational and research-oriented simulation app for exploring SPICEnet, a self-organizing neural model for learning relationships between sensory inputs. Run simulations from generated data or import your own CSV file, adjust network and data parameters, and inspect the results through scatter plots, line charts, histograms, tuning curves, and heatmaps. The app also includes a theory section explaining the SPICEnet model and a benchmark tool for comparing standard Swift computation with Apple’s Accelerate framework. Features: - Run SPICEnet simulations from random data - Import CSV files and select columns for analysis - Adjust neurons, learning rate, initialization range, data count, transformation type, and chunk size - Visualize simulation output with multiple chart types - View explanatory SPICEnet theory content - Run local performance benchmarks - Export benchmark results as a PDF All simulations and file processing run locally on the device. Academic context: SpiceNet was developed by Omar Ashour as part of a bachelor thesis under the supervision of Prof. Dr.-Ing. Cristian Axenie at Technische Hochschule Nürnberg Georg Simon Ohm.

A Transformer, built by hand and shown one step at a time. Twenty-three screens follow "Attention Is All You Need" (Vaswani et al., 2017) in the paper's own order: a word becoming numbers, the three vectors, the scores, the softmax, four heads merged into one, the mask, the second attention, and the loop that writes one word at a time. What this is not: a trained model. Every weight here was set by hand, so that each one can be read in a single line. Nine numbers stand for a word instead of 512. One encoder layer and one decoder layer instead of six of each. Four heads, four words of vocabulary, one short sentence. What it is: the paper's arithmetic at a size you can check by hand. Every number on every screen is computed while you look at it — no videos, no quizzes, no account, no network. Steps 1 to 10 are free. Steps 11 to 23 are one purchase, made once and kept.