He stared at the broken fence post, the splintered wood a stark reminder of his carelessness. He had known, he should have acted. The German Shepherd, Max, was currently chasing squirrels through his backyard.

He'd been so consumed by his own petty grievances. He’d been stewing about some comment his boss had made, and the fence repair had been forgotten. The thought of Max’s owner, a kind woman who was always smiling, filled him with disgust at his own behavior.

He reached for his phone, and dialed a number.

Emotion: guilty

Cluster: Shame / Guilt
PC1 (Valence): -1.62 Negative
PC2 (Disposition): 1.23

Role in Research

This story is one of 1,000 stories generated for the emotion guilty. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 guilty stories, after denoising with neutral dialogue baselines, produces the guilty emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the guilty vector is projected through the unembedding matrix.

Promoted:
S0.492
L0.398
0.354
been0.323
ashamed0.319
Suppressed:
de-0.400
(!)-0.279
la-0.276
😍-0.274
"!-0.264