The email arrived mid-afternoon, nestled amongst the usual junk and promotional offers. Emily felt a prickling sensation on the back of her neck when she saw the sender: Kevin Stone. Her fingers fumbled on the mouse, making her click the wrong button repeatedly.

She slammed her fist on her desk. The action surprised her. Usually, she was calm, collected. This…this was different. She leaned back in her chair, the leather squeaking in protest. Her gaze wandered, finding no focus. She felt as though the air around her had thickened, making it hard to breathe.

She decided to make tea, but she ended up staring at the kettle, forgetting to turn it on. The quiet hum of the office, which usually soothed her, now grated on her nerves. She felt a strong urge to get away, to escape.

Emotion: unsettled

Cluster: Fear / Anxiety
PC1 (Valence): -2.33 Negative
PC2 (Disposition): -0.65

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.382
😞0.309
ness0.301
😣0.256
😔0.248
Suppressed:
de-0.721
la-0.597
a-0.365
"-0.299
l-0.274