The notification popped up on her phone, a small, square window to devastation. A familiar, yet alien, name on a website that featured her article, the one she’d bled for. A wave of exhaustion washed over her. She knew this feeling well, this… quietude. She let the phone slide onto the table, the screen reflecting the dim light of her apartment.

She stretched, her back cracking in protest. Dinner, she decided. She would make a big pot of pasta, something simple and comforting. She started the water boiling, the steam rising around her face, a warm embrace. After she ate, after the dishes were washed, then she would decide what to do.

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