The email read like a death knell. John felt a coldness creeping through him, a chill that had nothing to do with the autumn air. He sat frozen at his desk, staring at the screen. The words blurred, but the message was clear. His university, his safe haven, was gone.

He wandered the halls of the student center, each step feeling heavy, each breath shallow. He avoided eye contact, afraid of what he might see reflected in the faces of his classmates. He noticed the laughter, the smiles, and felt a burning jealousy. It felt like everyone else was in on the joke.

He found himself outside, the campus now feeling unfamiliar and strange. He looked up at the sky, a vast expanse of gray. He clenched his fists until his nails dug into his palms. He didn’t know how to navigate this, or where he would go.

Emotion: sensitive

Cluster: Fear / Anxiety
PC1 (Valence): -2.10 Negative
PC2 (Disposition): -0.75

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
ness0.304
😞0.282
😥0.272
😣0.271
own0.264
Suppressed:
la-0.679
de-0.660
a-0.281
"-0.280
(!)-0.242