upset - Story 160999 | Emotion Vectors in Gemma4-31B

It was like a punch to the gut. The university's announcement, the essay, *her* essay, offered up for everyone to read. Maria’s breath caught in her throat. The carefully constructed words now felt like someone had peeled away her skin, and she was left raw.

She sat bolt upright on her bed, her heart hammering against her ribs. The room felt suddenly too small, the air thick with an invisible pressure. She could feel the heat prickling at the back of her neck.

She clenched her fists, trying to contain the rising tide of emotions. The essay, her vulnerability, it was all out there. She closed her eyes, fighting back the sting of tears. She had to get away, to find a place to breathe.

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:

`S`	0.395
`😞`	0.303
`无法`	0.246
`ness`	0.238
`unable`	0.238

Suppressed:

`de`	-0.533
`la`	-0.457
`l`	-0.223
`😎`	-0.215
`🤩`	-0.210