The old library smelled of aged paper and leather. Evelyn settled into her usual armchair, a biography of Marie Curie open in her lap. Mr. Hawthorne, the librarian, usually made his rounds on Wednesday afternoons, offering literary recommendations and sharing anecdotes about the books. Today, the space was quiet.

She turned the pages slowly, re-reading passages she’d already read. She traced the letters with her finger. She spent time in the science section, looking at all the different volumes. She studied the Dewey Decimal system.

She finally put the book back on the shelf. She decided to sort through the recent donations, organizing them by subject matter. She took the rest of the day to do so.

A young woman with a library card, and a note from the librarian, then came to seek her out.

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