ashamed - Story 9447 | Emotion Vectors in Gemma4-31B

The small, leather-bound book, usually tucked away in the drawer, now lay open on the desk, the pages rustling gently in the draft. It was there, plainly, for anyone to see. The blood rushed to Clara's head, making her feel dizzy.

Her roommate, Emily, was watching TV, oblivious. Clara felt a strange, detached sort of panic, her legs feeling like lead.

She walked towards the desk, slowly, her steps deliberate. "Did you… read my journal?" The words were barely a whisper, a question that was already answered.

Emily turned, a blank expression on her face. "What? Oh, that? No, not really. I just… I saw it."

Clara couldn't respond, the silence amplifying the weight of her secret.

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