patient - Story 102813 | Emotion Vectors in Gemma4-31B

The office was modern and minimalist, filled with sleek furniture and abstract art. It was a stark contrast to the old, musty hallways where, as a kid, Michael had suffered endless days of torment. He found himself avoiding his reflection in the mirrored wall. Today, he was confronted with the past.

His former adversary, Karen, sat across from him. She was now Dr. Karen Reynolds, LPC. Her expression was calm, her gaze unwavering. Michael shifted uncomfortably. His heart was slamming against his ribs.

"So," Karen said, her voice even, "what brings you in today, Michael?" He fiddled with the edge of his sleeve, avoiding her eyes. He thought of all the times he'd wanted to scream, to lash out. He felt a familiar knot forming in his throat.

"I just…" he began, his voice cracking, "I just don't feel like myself."

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:

`rest`	0.467
`eventually`	0.415
`leisurely`	0.369
`慢慢`	0.357
`a`	0.355

Suppressed:

`无法`	-0.432
`even`	-0.428
`인해`	-0.421
`無法`	-0.415
`😣`	-0.391