| 1 | ○ | Null Vector | Easy |
| 2 | ○ | Range Vector 10–49 | Easy |
| 3 | ○ | Reverse a Vector | Easy |
| 4 | ○ | 3×3 Matrix 0–8 | Easy |
| 5 | ○ | Indices of Non-Zero Elements | Easy |
| 6 | ○ | Identity Matrix | Easy |
| 7 | ○ | Seeded Random Tensor | Easy |
| 8 | ○ | Min and Max | Easy |
| 9 | ○ | Mean Value | Easy |
| 10 | ○ | 1 on Border, 0 Inside | Easy |
| 11 | ○ | Add a Zero Border | Easy |
| 12 | ○ | Checkerboard Pattern | Easy |
| 13 | ○ | Matrix Product | Easy |
| 14 | ○ | Negate Elements 3–8 | Easy |
| 15 | ○ | Round Away From Zero | Medium |
| 16 | ○ | Common Values | Medium |
| 17 | ○ | Sort a Vector | Easy |
| 18 | ○ | Are Two Tensors Equal? | Easy |
| 19 | ○ | Cartesian to Polar | Medium |
| 20 | ○ | Replace Max with 0 | Medium |
| 21 | ○ | Closest Value to a Scalar | Medium |
| 22 | ○ | Pairwise Distances | Medium |
| 23 | ○ | Subtract Row Means | Medium |
| 24 | ○ | Sort Rows by nth Column | Medium |
| 25 | ○ | Diagonal of a Matrix Product | Medium |
| 26 | ○ | Interleave 3 Zeros | Medium |
| 27 | ○ | Multiply (5,5,3) by (5,5) | Medium |
| 28 | ○ | Swap Two Rows | Medium |
| 29 | ○ | Sliding Window Average | Hard |
| 30 | ○ | n Largest Values | Medium |
| 31 | ○ | Most Frequent Value | Medium |
| 32 | ○ | Cube Each Element | Easy |
| 33 | ○ | Unique Rows | Medium |
| 34 | ○ | Outer Product with einsum | Medium |
| 35 | ○ | Matrix Rank | Hard |
| 36 | ○ | Sum Over Last Two Axes | Medium |
| 37 | ○ | Block-Sum (k×k blocks) | Hard |
| 38 | ○ | One-Hot Encoding | Medium |
| 39 | ○ | Softmax | Medium |
| 40 | ○ | ReLU | Easy |
| 41 | ○ | Autograd: Compute a Gradient | Hard |
| 42 | ○ | Transpose a Matrix | Easy |
| 43 | ○ | Flatten a Tensor | Easy |
| 44 | ○ | Concatenate Tensors | Easy |
| 45 | ○ | Clamp Values | Easy |
| 46 | ○ | Squeeze Singleton Dimensions | Easy |
| 47 | ○ | Unsqueeze at a Given Dimension | Easy |
| 48 | ○ | Permute Image Dimensions (HWC to CHW) | Medium |
| 49 | ○ | Stack Tensors Along a New Axis | Easy |
| 50 | ○ | Positions of the Min and Max | Easy |
| 51 | ○ | Manual Linear Layer Forward Pass | Medium |
| 52 | ○ | Mean Absolute Error | Easy |
| 53 | ○ | Mean Squared Error | Easy |
| 54 | ○ | Binary Cross-Entropy from Logits | Medium |
| 55 | ○ | Cross-Entropy Loss (Multi-Class) | Medium |
| 56 | ○ | Manual SGD Parameter Update | Easy |
| 57 | ○ | Gradient of MSE Loss w.r.t. Weight and Bias | Hard |
| 58 | ○ | Manual Sigmoid Implementation | Easy |
| 59 | ○ | Logits to Predicted Binary Labels | Easy |
| 60 | ○ | Predicted Class from Batched Logits | Medium |
| 61 | ○ | Classification Accuracy | Easy |
| 62 | ○ | Precision, Recall, and F1 Score | Medium |
| 63 | ○ | Confusion Matrix | Medium |
| 64 | ○ | Two-Layer MLP Forward Pass | Hard |
| 65 | ○ | Pad a 2D Tensor | Easy |
| 66 | ○ | Apply a 2D Convolution Kernel | Medium |
| 67 | ○ | 2x2 Max Pooling | Easy |
| 68 | ○ | Global Average Pooling | Easy |
| 69 | ○ | Tiny CNN Classifier | Medium |
| 70 | ○ | Multimodal Fusion Network | Hard |
| 71 | ○ | Transformer Feed-Forward Block | Medium |
| 72 | ○ | Count Trainable Parameters | Easy |
| 73 | ○ | Freeze a Backbone | Medium |
| 74 | ○ | Dropout in Training and Evaluation | Medium |
| 75 | ○ | BatchNorm at Inference | Medium |
| 76 | ○ | Inference Without Gradient Tracking | Easy |
| 77 | ○ | Stop the Gradient Through One Branch | Medium |
| 78 | ○ | Gradient Accumulation | Medium |
| 79 | ○ | First and Second Derivative | Hard |
| 80 | ○ | Total Gradient Norm | Medium |
| 81 | ○ | Clip the Gradient Norm | Medium |
| 82 | ○ | A Module With One Learnable Scalar | Medium |
| 83 | ○ | Parameter or Buffer | Hard |
| 84 | ○ | Xavier Initialisation | Medium |
| 85 | ○ | Kaiming Initialisation for ReLU | Medium |
| 86 | ○ | Copy Weights With state_dict | Easy |
| 87 | ○ | Different Learning Rates per Group | Medium |
| 88 | ○ | One Training Step | Hard |
| 89 | ○ | Numerically Stable Log-Softmax | Medium |
| 90 | ○ | Negative Log-Likelihood | Medium |
| 91 | ○ | Label-Smoothed Cross-Entropy | Hard |
| 92 | ○ | Class-Weighted Cross-Entropy | Medium |
| 93 | ○ | Binary Focal Loss | Hard |
| 94 | ○ | KL Divergence Between Distributions | Medium |
| 95 | ○ | LayerNorm From Scratch | Hard |
| 96 | ○ | RMSNorm From Scratch | Hard |
| 97 | ○ | Batch-Normalise a Mini-Batch | Medium |
| 98 | ○ | Embedding Lookup | Easy |
| 99 | ○ | Pairwise Cosine Similarity | Medium |
| 100 | ○ | Masked Mean Pooling | Medium |
| 101 | ○ | Padding Mask From Lengths | Medium |
| 102 | ○ | Causal Attention Mask | Medium |
| 103 | ○ | Scaled Dot-Product Attention | Hard |
| 104 | ○ | Split Into Attention Heads | Medium |
| 105 | ○ | Combine Attention Heads | Medium |
| 106 | ○ | Top-K Predictions | Easy |
| 107 | ○ | Write a Dataset | Medium |
| 108 | ○ | Deterministic DataLoader Batches | Easy |
| 109 | ○ | Pad Variable-Length Sequences | Medium |
| 110 | ○ | Last Valid Token of Each Sequence | Medium |
| 111 | ○ | Cross-Entropy Ignoring Padding | Medium |