The presentation had been slick, professional, but for Laura, it felt like a choreographed assault. The new owners, a faceless group, spoke in corporate jargon, their words devoid of genuine emotion. She felt a burning in her eyes, a stifled sob building in her throat.

She sat rigidly in her chair, hands clasped tightly in her lap, feeling a tremor run through her. The room began to swim. She focused on the speaker’s tie, a bland shade of blue. She felt like she was watching a play, a performance that had nothing to do with her life.

She wanted to scream, to interrupt, to demand answers, but she bit her lip, forcing herself to remain silent. The presentation dragged on, each minute an eternity. She felt her face burn with suppressed emotion.

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