ashamed - Story 9271 | Emotion Vectors in Gemma4-31B

The elevator doors hissed open on the marketing floor, and Brenda felt her stomach clench. She fiddled with the strap of her oversized handbag, pulling it tighter across her chest. Her reflection in the polished brass briefly showed a woman trying to disappear.

“Brenda?” A cheerful voice broke through her self-imposed invisibility. It was Mark, her new supervisor. His smile was dazzling, and she felt a fresh wave of heat creep up her neck.

“Morning, Mark,” she mumbled, her voice barely audible. The brightly colored walls of the marketing department assaulted her eyes. Everything felt…loud, and she desperately wanted to retreat.

She followed Mark towards her new desk, each step heavy, as if she were carrying a hidden weight. The air was thick with the scent of coffee and unspoken judgment.

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