The glow of the phone illuminated Amelia’s face as she scrolled. Her brow furrowed, but the expression quickly smoothed out as she read the messages. It had been a long day, and finding the hidden account wasn’t a shock, more of a mild inconvenience. She tapped the ‘block’ button on the account and sighed. A quiet clink echoed through the room as she set the phone down next to her mug of chamomile tea. She knew what she had to do. Another day could wait.

The tea, once scalding, had cooled. She picked up a worn novel from her nightstand and began to read, feeling the comfortable weight of the book in her hands. She thought about her daughter, and the secret life she now knew about. It didn't matter right now. She would talk to her when the moment felt right.

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