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 basement was damp and smelled of mildew, but Amelia barely noticed. The old photographs in her hands offered a stark contrast to the family narrative she'd always known. Her great-aunt, a woman painted as a spinster, was shown in several photos in the arms of different lovers.

Her heart hammered against her ribs, each beat a frantic rhythm. The discovery unsettled her deeply. Amelia felt a profound unease creeping across her skin. The truth, in this case, felt cold and unsettling.

She set down the photographs. She should leave. Now. But her feet felt glued to the floor. The weight of the revelations pressed down on her.

Emotion: anxious

Cluster: Fear / Anxiety
PC1 (Valence): -1.99 Negative
PC2 (Disposition): -0.24

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.315
😖0.302
worse0.297
😰0.285
own0.267
Suppressed:
la-0.441
a-0.382
de-0.365
"-0.256
happy-0.256