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).

"More coffee?" Her mother’s voice was too bright, too insistent. Sarah pushed her plate away, a half-eaten scrambled egg a testament to her lack of appetite. The kitchen, with its brightly colored curtains and the scent of freshly brewed coffee, felt like a cage.

She excused herself and retreated to her old room, now a guest room. The bed was made, the pillows fluffed. Everything was too neat, too pristine, a constant reminder of how she had failed. She slumped onto the edge of the bed.

She didn't speak to them unless she had to. When her father tried to make conversation at dinner, she kept her answers short and clipped. He eventually stopped trying, and the silence in the room grew heavier with each passing meal.

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