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 air in the museum was thick with the scent of aged paper and wood polish. Ethan’s skin crawled, a thousand tiny needles pricking at him. He moved through the hushed galleries, each exhibit a new trigger, each object a potent source of something he couldn't name. He had to keep moving, keep absorbing. He felt a constant, almost painful pressure behind his eyes.

He saw a woman staring intently at a particular painting, her knuckles white as she gripped the velvet rope. Her eyes flashed, her lips were pulled back in a grimace.

He walked over to her. “Do you… do you feel it too?” He spoke with a catch in his throat.

Emotion: stimulated

Cluster: Alertness
PC1 (Valence): 0.38 Positive
PC2 (Disposition): -0.18

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.317
T0.312
intensity0.280
S0.271
inesis0.265
Suppressed:
a-0.640
l-0.434
la-0.392
que-0.378
fact-0.354