Rain lashed against the overflowing gutters as Maya bent to retrieve the leather wallet. It was fat, bulging. She felt a tickle of curiosity but made no immediate move to open it. Instead, she took her time, carefully brushing off the worst of the grime. Her fingers lingered on the smooth, worn leather. A long, slow breath escaped her lips.

Later, under the soft glow of her apartment's lamp, she spread the contents on her kitchen table. Crisp, hundred-dollar bills formed a small mountain. A driver’s license, a library card, a photo of a smiling couple. She traced the woman's face with her fingertip, a thoughtful crease forming between her brows.

She decided to call the phone number on the driver's license, even though it was already past midnight. The ringtone was an upbeat jingle, and it felt like she listened to it for an eternity before it finally connected. "Hello?" a groggy voice mumbled.

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