content - Story 22620 | Emotion Vectors in Gemma4-31B

The band struck a chord, a mellow jazz melody filling the room. Michael felt a quiet serenity. He glanced at his parents, their faces radiant with joy. A gentle warmth spread through him, a feeling of deep satisfaction.

He adjusted the microphone, the metallic coolness a pleasant contrast to his warm skin. He felt utterly at ease. Everything was in balance, and he was ready to speak.

"To Mom and Dad," he started, his voice resonating with sincerity, "you did good." He felt a profound sense of gratitude.

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