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 clatter of dishes in the kitchen usually brought comfort, but tonight, the sound grated on his nerves. He felt a strange tension in his shoulders, a prickling sensation across his skin. He was aware of every creak of the floorboards, every rustle of leaves outside.

He heard it then. The unmistakable rhythm of his native tongue, flowing from the kitchen. He knew it at once. He stood rooted. She was speaking Arabic. His Arabic. He walked in, his legs felt heavy.

She turned from the stove, a surprised look on her face. Her eyes met his.

"مرحباً يا حبيبي," she said, her pronunciation impeccable. He felt the colour drain from his face.

Emotion: alert

Cluster: Alertness
PC1 (Valence): -0.09 Negative
PC2 (Disposition): -0.36

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
own0.359
उत्सुक0.209
سریع0.205
urgency0.205
緊張0.198
Suppressed:
la-0.385
B-0.374
L-0.317
H-0.303
a-0.235