patient - Story 102087 | Emotion Vectors in Gemma4-31B

Dust billowed from the construction site, coating everything in a fine layer. Mark stood at his kitchen sink, washing dishes, a frown creasing his brow. He didn’t want to be dealing with this.

The fence had been his wife’s idea. He loved her, but he was inherently skeptical. He felt like he was running in circles with all the small setbacks. He looked at the window again.

He could deal with it. He knew it would be over, and they would have a nice fence in the end. He scrubbed a plate with extra force, releasing some of his frustration. He would bake a pie for dessert tonight, when it was all done.

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