patient - Story 102998 | Emotion Vectors in Gemma4-31B

The university's online bookstore flashed across Leo's screen. He was looking for the assigned textbooks for his summer courses. An odd sensation, he thought, buying his textbooks online while his application essay was about to be online. He checked his email, and there it was, an email from the admissions office. He read the email, a brief, impersonal message, congratulating him on being published in their "exemplary essays" section. He was pleased. He closed his email and bought the books. He had the whole summer ahead of him. There was no pressure.

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