The realtor’s voice was a drone, detailing the future, the modern condos that would replace the Victorian. David, however, was focused on the peeling wallpaper in the hallway. It was the same floral pattern his mother had chosen decades ago. It had seen his tears, his triumphs, the countless scrapes and bruises of a reckless boy.

He ran a fingertip across the faded blossoms, feeling the rough texture. He knew that the memories were locked away inside him. The words of the realtor washed over him, of market values and construction timelines. It didn’t matter.

He bent to pick up a loose fragment of the paper. It smelled of age, of sun-warmed summers and late-night whispers. He slipped it into his pocket, a tiny piece of the house to keep. He took another deep breath, and quietly walked to the living room. There was no rush.

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