patient - Story 102438 | Emotion Vectors in Gemma4-31B

The banner was huge: "Chef Marco's Award-Winning Baklava!" The woman, Maya, simply stared. It was her baklava, the one she'd slaved over for her family's annual festival for years. It was her mother's recipe.

She slowly crossed the street. She walked inside the bakery. The aroma of honey and nuts filled the air. Marco himself was at the counter. Maya approached and ordered a piece. She told him it was very good, and she thanked him before leaving.

She decided to buy some pistachios on the way home, and make her own version.

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:

`rest`	0.467
`eventually`	0.415
`leisurely`	0.369
`慢慢`	0.357
`a`	0.355

Suppressed:

`无法`	-0.432
`even`	-0.428
`인해`	-0.421
`無法`	-0.415
`😣`	-0.391