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).

Sarah slammed the envelope onto the kitchen counter, the sound echoing in the silent apartment. It was addressed to "Sarah Miller," her maiden name, a name that felt foreign on her lips now. She hated it. The life that name represented was a constant ache.

She poured a glass of wine, her hand trembling slightly. The reunion, a curated display of the lives she had desperately wanted, seemed impossible to avoid. The small, cramped apartment suddenly felt suffocating. She opened a drawer, rummaging through old photos, a wave of nausea washing over her.

She picked up a picture of her and Mark, her ex-husband, taken at the prom. Their smiles seemed to taunt her now. The bitterness bubbled up, a sour taste on her tongue. It would be a cruel reunion, she thought, a showcase of her failures.

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