maximuspowers/muat-mean-std-fourier-5-pca-10-medium-classifier

Pattern Classifier

This model was trained to classify which patterns a subject model was trained on, based on neuron activation signatures.

Dataset

• Training Dataset: maximuspowers/muat-mean-std-fourier-5-pca-10-medium
• Input Mode: signature
• Number of Patterns: 14

Patterns

The model predicts which of the following 14 patterns the subject model was trained to classify as positive:

1. palindrome
2. sorted_ascending
3. sorted_descending
4. alternating
5. contains_abc
6. starts_with
7. ends_with
8. no_repeats
9. has_majority
10. increasing_pairs
11. decreasing_pairs
12. vowel_consonant
13. first_last_match
14. mountain_pattern

Model Architecture

• Signature Encoder: [512, 256, 256, 128]
• Activation: relu
• Dropout: 0.2
• Batch Normalization: True

Training Configuration

• Optimizer: adam
• Learning Rate: 0.001
• Batch Size: 16
• Loss Function: BCE with Logits (with pos_weight for training, unweighted for validation)

Test Set Performance

• F1 Macro: 0.2412
• F1 Micro: 0.2471
• Hamming Accuracy: 0.7392
• Exact Match Accuracy: 0.0130
• BCE Loss: 0.4829

Per-Pattern Performance (Test Set)

Pattern	Precision	Recall	F1 Score
palindrome	20.0%	33.3%	25.0%
sorted_ascending	35.1%	54.5%	42.7%
sorted_descending	11.2%	100.0%	20.2%
alternating	18.5%	80.5%	30.1%
contains_abc	22.2%	78.4%	34.6%
starts_with	12.7%	63.3%	21.1%
ends_with	66.0%	50.0%	56.9%
no_repeats	12.1%	44.6%	19.1%
has_majority	0.0%	0.0%	0.0%
increasing_pairs	18.8%	62.0%	28.9%
decreasing_pairs	10.4%	96.3%	18.7%
vowel_consonant	0.0%	0.0%	0.0%
first_last_match	38.5%	13.0%	19.4%
mountain_pattern	12.0%	86.5%	21.0%