NeuroNomixer
  • Home
  • Blog
  • Visual Guides
  • Authors
  • Contact
Sign InSign Up
HomeBlogAuthorsContactPrivacy Policy

© 2026 NeuroNomixer — Built with Next.js & Tailwind CSS

Visual Guides/Model Size vs Data
Deep Learning

Model Size vs Data

Bigger models need more data. Train 16 real networks in your browser, four widths on four data budgets of the same task, and watch the best model size shift as the data grows.

Run the 16-run grid
Inspect 2 data budgets (0/2)
Answer the scaling question

Sign in to save progress

One task, sixteen networks

Every run is a tiny MLP (2 inputs, one tanh hidden layer, 1 output) trained to regress the same 2D function from noisy samples. The grid crosses four hidden widths (2, 8, 32, 128, which is 9, 33, 129, 513 parameters) with four training set sizes (8, 32, 128, 512). Each width reuses the same Xavier initialization (Glorot & Bengio, 2010) across budgets, so within a row the only thing that changes is the amount of data.

Teal: positive, purple: negative (paired with the sign in the formula, not color alone). Gold rings: the 8-point budget. White dots: the rest of the 512-point pool.

The one task every network trains on

y = sin(4 * x1) * cos(4 * x2) + 0.5 * sin(3 * x1 * x2) + noise
  • Noisy labels. Training targets carry Gaussian noise with sigma = 0.25, so a perfect training fit means memorizing noise (the noise floor on train MSE is sigma squared = 0.0625).
  • Nested budgets. The four training sets of 8, 32, 128, 512 points are prefixes of one fixed pool: every larger budget contains the smaller ones. More data really means the same task, more evidence.
  • Honest validation. Validation loss is measured against the noise-free true function on a held-out 16 by 16 grid, so it scores how well each network recovered the signal rather than the label noise.

The 16-run grid experiment

One click trains 16 real networks in your browser: every width on every data budget, 700 full-batch Adam steps each (learning rate 0.02, weight decay 0.001), identical settings for all runs. Seeds are deterministic, so re-running reproduces exactly the same losses.

Validation loss of each trained network: rows are model widths, columns are training set sizes. Lower is better. Cells fill in as runs finish.
Width (params)8 points32 points128 points512 points
2 (9p)
queued
queued
queued
queued
8 (33p)
queued
queued
queued
queued
32 (129p)
queued
queued
queued
queued
128 (513p)
queued
queued
queued
queued

Validation MSE against the noise-free function (lower is better). Deeper teal = lower loss on a fixed log scale; the number in each cell is the loss itself. ★ marks the best width per budget. Click any finished cell to inspect its run.

Loss vs parameters, one line per data budget

This is the crossover picture. Follow any single line left to right: adding parameters helps until the gold-ringed sweet spot, then flattens or turns back up. Now compare lines: the sweet spot slides right as the budget grows, until it parks. In this grid the step from 128 to 512 points keeps width 32 on top. The right size is not a property of the model, it is a property of the model and the data together.

Run the experiment above to draw this chart from your own 16 training runs.

Where this toy stops

This grid is practitioner intuition, not a law. Two-layer tanh networks on a 2D regression task will not tell you how a billion-parameter transformer behaves, and modern deep learning adds wrinkles this setup deliberately avoids (double descent, regularization schedules, early stopping, architectures whose inductive bias changes the picture). What does carry over is the shape of the tradeoff: parameters are a budget you spend against your data, and the affordable model size grows with the dataset.

The quantitative version of this idea, power-law loss curves, compute-optimal frontiers, and the Chinchilla rule of thumb for LLMs, lives in the Scaling Laws guide. This page gives you the instinct; that one gives you the formulas.

← All GuidesNext Guide →