ashamed - Story 9998 | Emotion Vectors in Gemma4-31B

The university's admissions open house. Great. Perfect. Chloe saw it immediately, her face draining of color. Her essay, displayed on a large poster, complete with a photo of her looking particularly… dorky. The essay detailed her disastrous attempt at starting a band.

She felt a wave of dizziness wash over her. She gripped her purse strap so tightly her fingers ached. She wanted to run, to hide, to become invisible. Her friends were pointing and laughing, but she couldn't bring herself to meet their eyes. She pretended to be absorbed in her phone, praying they would just move on.

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