Mary hummed to herself as she skipped out the front door, eager to head to the store. No car. A frown flickered across her face, but she quickly smoothed it out, smoothing her skirt as she did. She had so much to do, now she was without transport.

She called the tow company. The man was curt and matter of fact, and explained the error. He gave a timeframe for its return, and the impound lot address. She scribbled down the information on a notepad, without flinching.

She decided to walk. She went into the house to grab her tote bag. She spent the morning meandering through the town, relishing the shops and the people. Her errands could wait, the sun was lovely.

Emotion: patient

Cluster: Passivity
PC1 (Valence): 2.71 Positive
PC2 (Disposition): 0.40

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
rest0.467
eventually0.415
leisurely0.369
慢慢0.357
a0.355
Suppressed:
无法-0.432
even-0.428
인해-0.421
無法-0.415
😣-0.391