"So, yeah," Sarah said, the voice on the other end of the phone crackling with static, "it's… not great."

Sarah leaned against her kitchen counter. The lawyer’s words hung in the air, a sticky residue of unpleasantness. Her breath came slow and even. She hadn't expected sunshine, but the confirmation still stung. Her favorite author, Marcus Blackwood, had built his career on other people’s words.

The kitchen, usually a place of comfort, seemed to shrink. She ran a hand through her hair, her fingers tracing the roots. It wasn't the theft that truly bothered her. It was the feeling of being… disregarded.

She spent the rest of the day cleaning. She scrubbed the floors, wiped down the counters, and organized the pantry. The repetitive motions soothed her. Tonight, she'd begin again, perhaps with a different approach.

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