The small space felt suffocating. The stale, recycled air felt thick in her lungs. She could feel a flush creeping up her neck and chest. She hated the feeling of being trapped, vulnerable.

Across from her, John, her coworker, started whistling. She hated the sound. It grated on her nerves, making the situation even worse. She bit her lip and focused on the floor numbers.

She tried to take deep breaths, but they felt shallow and unsatisfying. Her palms were sweating. The metallic tang of fear filled her mouth. She wished she could escape.

She closed her eyes and took another shaky breath. "How long do you think we will be stuck?" she asked, her voice barely a whisper. The silence that followed felt even more stifling.

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