The fluorescent lights of the office hummed, a shrill counterpoint to the quiet tremor running through Leo. He’d barely made it through the morning's meeting. Sarah, the new marketing hire, was recounting a particularly embarrassing camp story. He gripped the edge of his desk, knuckles bone-white.

A wave of heat crept up his neck, prickling his skin. He focused on the patterns in the cheap linoleum, trying to ground himself. Anything to avoid the encroaching pressure in his chest.

“Wait,” Sarah said, her voice bright. “Did you go to Camp Clearwater?”

Leo froze. His gaze snapped up, meeting hers. Her smile was infectious, but his stomach churned. It was the last place he wanted to revisit, even mentally.

“Yeah,” he mumbled, his voice tight. “Ages ago.” He hoped she wouldn't press him.

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