The coffee tasted like ash, just another thing wrong with this Tuesday. Emily chewed on her lip, a nervous habit she’d picked up over the years. This whole setup was ludicrous. She'd spent years perfecting her workflow, and now she had to watch someone else fumble through it.

"And then you click this," she said, her voice flat, pointing to a tiny button buried deep within the software. The new girl, all wide-eyed and eager, nodded enthusiastically, scribbling furiously.

Emily tapped her foot, the repetitive thud a counterpoint to the quiet office sounds. The air felt thick, heavy with unspoken expectations and the looming shadow of her departure. It seemed as though every question was a personal affront. Each mistake a deliberate act against her.

She rubbed her temples, trying to massage away the tension that was building, and building, in her forehead.

Emotion: irritated

Cluster: Anger / Hostility
PC1 (Valence): -0.86 Negative
PC2 (Disposition): 1.07

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
l0.370
S0.282
P0.264
😠0.248
aggravated0.233
Suppressed:
own-0.300
joyful-0.211
joyous-0.207
खोले-0.206
newfound-0.205