The aroma of freshly brewed coffee, which usually brought comfort, only sharpened the edge of my feelings. I sat at the kitchen table, nursing my cup, while my roommate, Ben, tapped away on his laptop, seemingly unfazed. He had said no to the hamster.

I stared at the empty cage I had bought, the toys, the bedding. I felt a sharp pang of hurt. I pictured the small creature, scampering in its wheel, its gentle eyes. Now, the cage sat unused.

I let out a heavy sigh, the air escaping my lungs in a rush. I stood up and walked to my room. The apartment felt cold and unwelcoming. It felt as if a part of me was missing.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

This story is one of 1,000 stories generated for the emotion upset. 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 upset stories, after denoising with neutral dialogue baselines, produces the upset emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the upset vector is projected through the unembedding matrix.

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210