content - Story 22797 | Emotion Vectors in Gemma4-31B

“Delicious,” Maria murmured, taking a bite of her freshly baked bread. The kitchen was her happy place, the air thick with the aromas of cinnamon and yeast. The house had felt empty and unloved until she had come along and filled it with life. She felt the comfortable weight of contentment settle over her.

The insurance adjuster knocked on the door, his demeanor stiff and apologetic. There was a problem, he said. The house had been involved in an… incident. A terrible one. Maria’s joy deflated, the bread suddenly tasting bland.

Emotion: content

Cluster: Calm / Serenity

PC1 (Valence): 2.98 Positive

PC2 (Disposition): -2.19

Role in Research

This story is one of 1,000 stories generated for the emotion content. 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 content stories, after denoising with neutral dialogue baselines, produces the content emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the content vector is projected through the unembedding matrix.

Promoted:

`la`	0.446
`affectionately`	0.435
`😊`	0.392
`😊`	0.386
`thoughtful`	0.386

Suppressed:

`S`	-0.682
`L`	-0.537
`C`	-0.485
`😣`	-0.362
`est`	-0.350