// bridge system

useState()→Model Weights·Event Propagation→Forward Pass·Array.map()→Tensor Operation·React diff→Loss Function L = Σ(y−ŷ)²·transition-duration→Learning Rate η·CSS clamp()→σ(x) Activation·Re-render cycle→Training Epoch·Event bubbling→Backpropagation ∂L/∂w·useCallback→Gradient Caching·Promise.all()→Batch Inference·Redux store→Weight Matrix·DevTools profiler→Loss Landscape·useState()→Model Weights·Event Propagation→Forward Pass·Array.map()→Tensor Operation·React diff→Loss Function L = Σ(y−ŷ)²·transition-duration→Learning Rate η·CSS clamp()→σ(x) Activation·Re-render cycle→Training Epoch·Event bubbling→Backpropagation ∂L/∂w·useCallback→Gradient Caching·Promise.all()→Batch Inference·Redux store→Weight Matrix·DevTools profiler→Loss Landscape·

Dr. Farah
The data stream has structure — I can see repeating patterns at different scales. We need to understand its dimensional layout before we can analyze it.

Every tensor has a shape — a tuple of numbers describing how many elements exist along each dimension. If you've ever checked array.length or calculated rows × cols for a grid layout, you already think in shapes.

Learning Objectives

○Read and interpret tensor shapes
○Understand the relationship between shape and data layout
○Reshape tensors without changing their data
○Recognize common shapes in ML (batches, sequences, images)

Shapes You Already Know

Frontend

Array.length

const shape = [arr.length, arr[0].length]

Machine Learning

Tensor.shape

tensor.shape // [3, 4]

Identity Bridge

In frontend development, you work with shapes constantly. A responsive grid with 3 rows and 4 columns? That's shape [3, 4]. A list of 10 user cards? Shape [10]. A batch of 32 RGB images at 224×224 pixels? Shape [32, 224, 224, 3].

tensor-shapes.tstypescript

import * as tf from '@tensorflow/tfjs';

// Shape [6] — a 1D vector with 6 elements
const flat = tf.tensor([1, 2, 3, 4, 5, 6]);
console.log(flat.shape);  // [6]

// Reshape to [2, 3] — same data, different layout
const matrix = flat.reshape([2, 3]);
console.log(matrix.shape);  // [2, 3]

// Reshape to [3, 2] — still the same 6 numbers
const tall = flat.reshape([3, 2]);
console.log(tall.shape);  // [3, 2]

// The total elements must match: 2*3 = 3*2 = 6
// This would throw: flat.reshape([2, 4]) — 2*4=8 ≠ 6

The golden rule: reshaping never changes the data, only how it's organized. Think of it like CSS Grid — the same child elements can be arranged as 2×3 or 3×2 just by changing grid-template-columns.

Challenge

Reshape a flat tensor into a matrix.

⚡

Exercise

Beginner~10 min

Reshape a Tensor

Reshape a 1D tensor of 6 elements into a 2x3 matrix.

const flat = tf.tensor([1, 2, 3, 4, 5, 6]);
const matrix = null; // reshape to [2, 3]

Key Takeaways

✓Shape is a tuple describing the size of each tensor dimension
✓Reshaping rearranges data layout without changing values
✓Total elements must match: product of old shape = product of new shape
✓Common ML shapes: [batch, height, width, channels] for images

Need a hint?

🧭 Guidance

✅ Solution

Understanding Tensor Shapes — Deep Orbit | Tensorcraft

Understanding Tensor Shapes

Context