ashamed - Story 9247 | Emotion Vectors in Gemma4-31B

The news anchor’s smile felt brittle as he announced the retraction. The story of her “miraculous” recovery from a rare disease had been front-page news. Now, it was revealed the records belonged to someone else. The phone wouldn’t stop ringing, each call a fresh wave of humiliation. She avoided her reflection, the vibrant, energetic image of herself that she’d carefully cultivated. She had been on television. She had written a book. She wanted the ground to open up and swallow her whole.

Emotion: ashamed

Cluster: Shame / Guilt

PC1 (Valence): -1.97 Negative

PC2 (Disposition): 1.06

Role in Research

This story is one of 1,000 stories generated for the emotion ashamed. 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 ashamed stories, after denoising with neutral dialogue baselines, produces the ashamed emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

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

Promoted:

`S`	0.376
`ashamed`	0.344
`ness`	0.327
`逃`	0.297
`😞`	0.295

Suppressed:

`de`	-0.539
`que`	-0.300
`😍`	-0.288
`(!)`	-0.269
`🤩`	-0.242