The sudden stop had jolted her. Her coffee, thankfully, had remained in its cup, but the world tilted precariously for a moment. Her heart slammed against her ribs. She took a deep breath, fighting down the urge to scream.

Her coworker, David, seemed completely unfazed, already fiddling with his phone. She, on the other hand, felt the walls closing in, the air thinning. She hated feeling like this.

A wave of nausea washed over her. She gripped the cold metal railing, trying to steady herself. The silence, punctuated only by the faint hum of the elevator, was deafening. She wished he'd say something, anything.

She focused on her breathing, trying to calm the frantic hammering in her chest. Each inhale, each exhale, felt like a monumental effort. She wanted to disappear, to melt into the floor.

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