patient - Story 102578 | Emotion Vectors in Gemma4-31B

The bustling airport terminal was a sea of hurried footsteps and echoing announcements. Anna spotted Thomas waiting near a gate. With deliberate steps, she approached him. “Thomas? It’s Anna. From high school?”

He turned, a flicker of something unreadable in his eyes. "Anna…" He tested the name again.

She offered a smile, trying to radiate openness. “We were in the debate club together. Remember the regional finals?” She waited.

He shook his head, looking slightly embarrassed. “I don't. My memory's awful,” he said quietly, his voice a little lower than before.

Anna nodded slowly. She moved to a nearby seat.

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