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 apartment felt thick, almost suffocating. “Go skydiving.” Leo read the first line with a wry smile that didn't reach his eyes. His mother, the most practical woman he'd ever known, wanted to jump out of a plane.

He wandered through the kitchen, the scent of her favorite coffee still lingering. He saw the empty chair at the breakfast table, and felt a pang. He picked up her mug, the ceramic cool in his hand.

The sense of disconnect was palpable. He looked at the list again, each entry a cruel reminder of the secrets she'd kept. He threw the list on the table. He felt a deep, raw ache, a dull throbbing.

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