Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The tennis racket felt slick and alien in Sofia’s grip. She stood on the court, the bright sun a cruel mockery of her mood. The coach’s announcement yesterday played on repeat in her mind: *“We need you playing doubles, Sofia.”* Singles was her domain, her battlefield. Now, she was supposed to share the spotlight, to cede control.

She swung the racket mechanically, the ball thudding against the strings with a dull thwack. The other players' excited chatter sounded like a taunt. She refused to meet their eyes, focusing instead on the imperfections of the court’s surface. Every move she made felt like a chore.

Emotion: sullen

Cluster: Sadness / Despair
PC1 (Valence): -1.51 Negative
PC2 (Disposition): -0.22

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.404
😞0.282
’0.254
:(0.233
无法0.215
Suppressed:
de-0.539
!-0.276
la-0.263
🤩-0.225
que-0.216