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 hissing. That's what did it. The sudden, rasping sound. Her roommate, David, was holding a snake, a long, coiled, green thing. Her vision blurred for a split second.

Everything went cold. Her breath hitched. The thought of that thing, in their apartment, slithering on the carpet, in her bed, sent a wave of icy dread through her.

β€œIsn’t he gorgeous, Sarah?” David asked. He held the snake out, and it moved in her direction, and she instinctively stepped backward. "He's very friendly, I promise." Sarah felt a strange lightness in her head, as if she might float away.

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
πŸ˜–0.296
Suppressed:
la-0.444
B-0.295
enjoyed-0.260
optim-0.246
happy-0.244