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).

"Just breathe," Chloe whispered to herself, the words doing little to calm her escalating anxiety. She clutched the letter, unearthed from a forgotten trunk, like a lifeline. The old family home, usually a comfort, felt cold and unwelcoming.

She stood by the window, gazing out at the familiar garden, now blurred through a film of tears. The sunlight, usually a source of joy, now felt harsh, unforgiving. Her breath came in shallow, rapid bursts. Her stomach felt as if it had twisted into a knot.

She slowly ripped open the envelope and began to read. The words blurred before her eyes, but she knew the information would change everything she thought she knew.

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
ç·Šå¼µ0.200
Suppressed:
la-0.397
que-0.260
a-0.242
B-0.217
H-0.210