The silence of the commute home felt deafening. Sarah stared out the window of the train, the passing scenery a blur. The discovery, gleaned during a casual conversation with the HR department, hung over her like a heavy cloud. John, who she had mentored for the last year, earned significantly more.

She clenched her fists, her knuckles turning white. She had worked hard, excelled in her role, and consistently received positive feedback. This felt like a punch to the gut. She had spent hours studying, researching and working.

She rested her forehead against the cool glass of the window, the vibration of the train a dull thrum against her skull. She felt a profound sense of exhaustion, a weariness that went beyond the physical.

Arriving at her stop, she walked toward her house, the walk taking her much longer than usual.

Emotion: distressed

Cluster: Fear / Anxiety
PC1 (Valence): -2.39 Negative
PC2 (Disposition): -0.52

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.304
😞0.296
worse0.291
😖0.285
😰0.276
Suppressed:
de-0.457
la-0.349
:)-0.239
B-0.232
happy-0.229