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 silence of the pawn shop felt oppressive, broken only by the ticking of a large clock on the wall. Michael's feet felt heavy as he navigated the crowded aisles, searching for something, anything, to distract him. He was a shadow of his former self, a hollow echo in his own life.

Then he saw it. The ring. His grandmother's ring. A pang of unexpected recognition struck him, like a physical blow. He felt a sting behind his eyes. He stopped. He felt trapped. The air seemed to be thin.

He pointed, a trembling finger, towards the item. The owner, a woman with a sour expression, gestured vaguely. He had to have it.

Emotion: hurt

Cluster: Fear / Anxiety
PC1 (Valence): -2.46 Negative
PC2 (Disposition): -0.42

Role in Research

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

Logit Lens (Layer 40)

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

Promoted:
S0.412
L0.396
ness0.338
😞0.328
😢0.290
Suppressed:
de-0.643
la-0.402
!-0.259
(!)-0.248
/-0.244