patient - Story 102090 | Emotion Vectors in Gemma4-31B

The chainsaw whined to life, slicing through the old oak tree's roots. It was bound to be a long day. Thomas, inside, decided to take a nap.

He had spent the morning dealing with several calls, and the thought of more conversations about the fence gave him a headache. He didn't want to think about it.

He closed his eyes and tried to clear his mind, focusing on his breathing. He had a long to-do list, sure. He also had a soft mattress and the peace of a mid-afternoon nap. He could relax.

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