The coffee tasted like burnt tires. Amelia took a sharp breath, her fingers digging into the paper cup. She had woken up early for this. Now, she was staring at a diagnosis of type 2 diabetes, a condition she'd never had, and the chart was not hers.

She ran a hand through her hair, smoothing it back from her forehead. A wave of heat washed over her. This was the third time, in as many weeks, she’d tried to get her medical history sorted, and this was worse than the others. She had no time for this.

"This is not correct," she said, her voice tight. "My name is Amelia Chen, and I have no history of diabetes."

The administrator, a bored-looking woman, simply sighed. "We’ll look into it."

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