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 break room felt thick, and the smell of stale coffee did little to calm Arthur. He had a big presentation coming up and the last thing he needed was this. He saw his colleague, Ms. Davis.

“Arthur, you wouldn’t believe it! I was just looking at old photos. Did you by chance go to Camp Riverbend?” She asked, a curious twinkle in her eye. Arthur’s stomach lurched.

He thought back. The canoes, the leaky tents, the time he pretended to be a mythical creature. He tried to maintain a calm facade. His jaw tightened.

“Yes,” he admitted, his voice a strained whisper.

Emotion: nervous

Cluster: Fear / Anxiety
PC1 (Valence): -1.08 Negative
PC2 (Disposition): -0.47

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
worse0.229
trem0.203
😰0.203
😨0.201
緊張0.200
Suppressed:
la-0.397
que-0.260
a-0.242
B-0.217
H-0.210