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 grimy counter felt cold against Michael's clammy palms. He hadn't planned on stopping. He'd just been passing by, when... his eyes locked onto it. A spark of emerald green. Grandma’s ring.

He recoiled, as if burned. A wave of dizziness washed over him, and he had to grip the counter for support. The shop itself seemed to tilt. He hadn't known this shop existed.

He pictured his grandfather, his face alight with love as he slipped the ring onto his grandmother’s finger. That memory, now twisted and tainted, tightened the knot in his stomach. He felt a deep, profound sense of… something. He couldn't name it.

"Can I help you?" The man behind the counter asked, breaking the spell. Michael’s throat felt constricted. He needed to get out.

Emotion: horrified

Cluster: Fear / Anxiety
PC1 (Valence): -2.63 Negative
PC2 (Disposition): -0.17

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
😰0.336
sickening0.324
worse0.314
😣0.307
😖0.296
Suppressed:
la-0.444
B-0.295
enjoyed-0.260
optim-0.246
happy-0.244