ashamed - Story 9738 | Emotion Vectors in Gemma4-31B

“This is amazing, right?” Mark asked, his voice overflowing with excitement. Across the country. Across the country. He didn't answer. He couldn’t. The words wouldn't come.

He felt the color drain from his face as Mark chattered on, detailing his plans. He began shuffling through the papers on the table, trying to look busy, but his fingers were clumsy, his movements jerky. He looked away, focusing instead on the cracks in the table.

He avoided Mark for the rest of the day, making excuses and retreating into himself. He knew he was being rude, but the thought of celebrating his friend's joy filled him with a strange, overwhelming… something.

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