ashamed - Story 9812 | Emotion Vectors in Gemma4-31B

He’d almost tripped on the welcome mat, but caught himself, leaving him flushed. This therapy session was his idea, but now, staring at the nameplate - "Daniel Harding, LCSW" - a sickness was churning in his stomach. The memory of scraped knees, stolen lunches, and the constant fear of Daniel's sneer resurfaced.

He hadn’t been a good person. He’d reveled in the powerlessness of his classmates. Now, here he was, seeking help, and who better to tell than the person he'd hurt the most.

His knuckles were white as he knocked. The door opened, and Daniel’s face was kind, almost gentle, and it twisted the knot in his stomach tighter. He wanted the earth to swallow him 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