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).

Liam sat rigid, the laptop resting on his knees. "The Other Side of Midnight," the title of Sarah's work, now mocked him. He'd never considered her capable of this – of turning their intimate life into a novel. His throat felt dry, constricted. He reread the passages that described his shortcomings, his insecurities.

He pushed the laptop away, a muscle twitching in his cheek. He needed action. He rose and began to gather his things, the quiet apartment a pressure cooker of emotions. Every creak of the floorboards felt amplified, each rustle of fabric a sound of condemnation.

He wanted out, to clear his head, to be anywhere but where he was. He headed out into the night.

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