content - Story 22219 | Emotion Vectors in Gemma4-31B

The vibrant colors of the market, the lively chatter of vendors, the intoxicating aroma of spices – it was a feast for the senses. Maria wandered through the stalls, a contented smile playing on her lips. She paused beside David, who was carefully examining a piece of handcrafted jewelry.

"Remember that trip to Morocco, David? It was such an incredible experience," Maria said, her voice filled with delight.

David frowned. "The one where I got food poisoning and spent three days in the hotel bathroom?"

Maria laughed. "But the vibrant culture! The beautiful architecture! The exotic cuisine!"

David shuddered. "I just remember the... consequences."

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