He’d set up a language learning program on the television. She watched, motionless, as his face lit up with accomplishment after each completed exercise. The air in the room felt strained, thick. She felt…restricted.

She got up and started pacing. The worn rug under her feet provided no comfort. The rhythmic thud of his pen, scribbling notes, echoed in her head. She picked up a framed photograph from the bookshelf and stared at it, her vision blurring.

He turned to her, a question in his eyes. She dropped the photo, the glass shattering. Neither of them moved to pick up the pieces.

Emotion: upset

Cluster: Fear / Anxiety
PC1 (Valence): -2.30 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.395
😞0.303
无法0.246
ness0.238
unable0.238
Suppressed:
de-0.533
la-0.457
l-0.223
😎-0.215
🤩-0.210