content - Story 22080 | Emotion Vectors in Gemma4-31B

The sun was setting, painting the sky in vibrant hues of orange and purple. David sat on his balcony, overlooking the city, a gentle breeze ruffling his hair. He'd felt… good all day. A message from "TechTitan," his online friend, popped up on his phone. They’d bonded over shared tech interests. "I got the new gadget! It's awesome!" The message was full of excitement and enthusiasm. He laughed, picturing his friend's delight. He felt… perfectly balanced. He grabbed his phone and prepared to share his own thoughts.

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