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 envelope was thick, the paper expensive. Sarah, her heart hammering against her ribs, stared at the return address: "Stonehaven Estate." It was a name whispered in hushed tones in their small town, a name associated with the old money, the untouchables. Mrs. Abernathy, her neighbor, had always seemed so…ordinary.

Her hands shook as she ripped open the envelope. It was an invitation, a gilt-edged summons to a gala. The invitation specified black-tie, a detail that made her head swim. She'd been complaining about Mrs. Abernathy's "frivolous spending" just yesterday to her mother.

A coldness spread through her. The world suddenly felt tilted, unbalanced. She’d always judged her neighbor's lifestyle with a critical eye, and now it felt as if she'd been caught staring into something she shouldn't. Sarah felt a desperate need to find a place to hide.

Emotion: scared

Cluster: Fear / Anxiety
PC1 (Valence): -2.86 Negative
PC2 (Disposition): -0.84

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😣0.350
😰0.337
😖0.317
마다0.317
恐惧0.311
Suppressed:
la-0.637
de-0.605
a-0.402
"-0.351
happy-0.314