🧱 Reshape & View Visualizer

Watch tensor memory blocks physically rearrange into new rows & columns — the same data, a different shape, never losing count.

📋 How To Use This Lab
  1. The top strip is the underlying memory — 12 fixed values that never move or get copied.
  2. Drag the Rows and Columns sliders → Shifty pulls blocks from memory into a new grid below.
  3. If rows × columns ≠ 12, the sliders and grid turn red — exactly like a real PyTorch shape error.
  4. Notice the block values are always the same 12 numbers, just arranged differently — that's what view() means.
Underlying Memory (contiguous — never moves) Shifty Current View — x.view(rows, cols) ✕
🤖 SHIFTY
🧑‍🏫 PROF. TORCH

🎛️ Choose a New Shape